Attention is required for canonical brain signature of prediction error despite early encoding of the stimuli.

Prediction error is a basic component of predictive-coding theory of brain processing. According to the theory, each stage of brain processing of sensory information generates a model of the current sensory input; subsequent input is compared against the model and only if there is a mismatch, a prediction error, is further processing performed. Recently, Smout and colleagues found that a signature of prediction error, the visual (v) mismatch negativity (MMN), for a fundamental property of visual input-its orientation-was absent without endogenous attention on the stimuli. This is remarkable because the weight of evidence for MMNs from audition and vision is that they occur without endogenous attention. To resolve this discrepancy, we conducted an experiment addressing 2 alternative explanations for Smout and colleagues' finding: that it was from a lack of reproducibility or that participants' visual systems did not encode the stimuli when attention was on something else. We conducted a similar experiment to that of Smout and colleagues. We showed 21 participants sequences of identically oriented Gabor patches, standards, and, unpredictably, otherwise identical, Gabor patches differing in orientation by ±15°, ±30°, and ±60°, deviants. To test whether participants encoded the orientation of the standards, we varied the number of standards preceding a deviant, allowing us to search for a decrease in activity with the number of repetitions of standards-repetition suppression. We diverted participants' attention from the oriented stimuli with a central, letter-detection task. We reproduced Smout and colleagues' finding of no vMMN without endogenous attention, strengthening their finding. We found that our participants showed repetition suppression: They did encode the stimuli preattentively. We also found early processing of deviants. We discuss various explanations why the earlier processing did not extend into the vMMN time window, including low precision of prediction.

Multicentre validation of CT grey-level co-occurrence matrix features for overall survival in primary oesophageal adenocarcinoma.

BACKGROUND: Personalising management of primary oesophageal adenocarcinoma requires better risk stratification. Lack of independent validation of proposed imaging biomarkers has hampered clinical translation. We aimed to prospectively validate previously identified prognostic grey-level co-occurrence matrix (GLCM) CT features for 3-year overall survival. METHODS: Following ethical approval, clinical and contrast-enhanced CT data were acquired from participants from five institutions. Data from three institutions were used for training and two for testing. Survival classifiers were modelled on prespecified variables ('Clinical' model: age, clinical T-stage, clinical N-stage; 'ClinVol' model: clinical features + CT tumour volume; 'ClinRad' model: ClinVol features + GLCM_Correlation and GLCM_Contrast). To reflect current clinical practice, baseline stage was also modelled as a univariate predictor ('Stage'). Discrimination was assessed by area under the receiver operating curve (AUC) analysis; calibration by Brier scores; and clinical relevance by thresholding risk scores to achieve 90% sensitivity for 3-year mortality. RESULTS: A total of 162 participants were included (144 male; median 67 years [IQR 59, 72]; training, 95 participants; testing, 67 participants). Median survival was 998 days [IQR 486, 1594]. The ClinRad model yielded the greatest test discrimination (AUC, 0.68 [95% CI 0.54, 0.81]) that outperformed Stage (ΔAUC, 0.12 [95% CI 0.01, 0.23]; p = .04). The Clinical and ClinVol models yielded comparable test discrimination (AUC, 0.66 [95% CI 0.51, 0.80] vs. 0.65 [95% CI 0.50, 0.79]; p > .05). Test sensitivity of 90% was achieved by ClinRad and Stage models only. CONCLUSIONS: Compared to Stage, multivariable models of prespecified clinical and radiomic variables yielded improved prediction of 3-year overall survival. CLINICAL RELEVANCE STATEMENT: Previously identified radiomic features are prognostic but may not substantially improve risk stratification on their own. KEY POINTS: • Better risk stratification is needed in primary oesophageal cancer to personalise management. • Previously identified CT features-GLCM_Correlation and GLCM_Contrast-contain incremental prognostic information to age and clinical stage. • Compared to staging, multivariable clinicoradiomic models improve discrimination of 3-year overall survival.

Predicting programmed death-ligand 1 (PD-L1) expression with fluorine-18 fluorodeoxyglucose ([18F]FDG) positron emission tomography/computed tomography (PET/CT) metabolic parameters in resectable non-small cell lung cancer.

BACKGROUND: Programmed death-ligand 1 (PD-L1) expression is a predictive biomarker for immunotherapy in non-small cell lung cancer (NSCLC). PD-L1 and glucose transporter 1 expression are closely associated, and studies demonstrate correlation of PD-L1 with glucose metabolism. AIM: The aim of this study was to investigate the association of fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography ([18F]FDG-PET/CT) metabolic parameters with PD-L1 expression in primary lung tumour and lymph node metastases in resected NSCLC. METHODS: We conducted a retrospective analysis of 210 patients with node-positive resectable stage IIB-IIIB NSCLC. PD-L1 tumour proportion score (TPS) was determined using the DAKO 22C3 immunohistochemical assay. Semi-automated techniques were used to analyse pre-operative [18F]FDG-PET/CT images to determine primary and nodal metabolic parameter scores (including max, mean, peak and peak adjusted for lean body mass standardised uptake values (SUV), metabolic tumour volume (MTV), total lesional glycolysis (TLG) and SUV heterogeneity index (HISUV)). RESULTS: Patients were predominantly male (57%), median age 70 years with non-squamous NSCLC (68%). A majority had negative primary tumour PD-L1 (TPS < 1%; 53%). Mean SUVmax, SUVmean, SUVpeak and SULpeak values were significantly higher (p < 0.05) in those with TPS ≥ 1% in primary tumour (n = 210) or lymph nodes (n = 91). However, ROC analysis demonstrated only moderate separability at the 1% PD-L1 TPS threshold (AUCs 0.58-0.73). There was no association of MTV, TLG and HISUV with PD-L1 TPS. CONCLUSION: This study demonstrated the association of SUV-based [18F]FDG-PET/CT metabolic parameters with PD-L1 expression in primary tumour or lymph node metastasis in resectable NSCLC, but with poor sensitivity and specificity for predicting PD-L1 positivity ≥ 1%. CLINICAL RELEVANCE STATEMENT: Whilst SUV-based fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography metabolic parameters may not predict programmed death-ligand 1 positivity ≥ 1% in the primary tumour and lymph nodes of resectable non-small cell lung cancer independently, there is a clear association which warrants further investigation in prospective studies. TRIAL REGISTRATION: Non-applicable KEY POINTS: • Programmed death-ligand 1 immunohistochemistry has a predictive role in non-small cell lung cancer immunotherapy; however, it is both heterogenous and dynamic. • SUV-based fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography ([18F]FDG-PET/CT) metabolic parameters were significantly higher in primary tumour or lymph node metastases with positive programmed death-ligand 1 expression. • These SUV-based parameters could potentially play an additive role along with other multi-modal biomarkers in selecting patients within a predictive nomogram.

AGA Clinical Practice Guideline on Systemic Therapy for Hepatocellular Carcinoma.

BACKGROUND & AIMS: Hepatocellular carcinoma (HCC), the most common primary liver cancer, remains a deadly cancer, with an incidence that has tripled in the United States since 1980. In recent years, new systemic therapies for HCC have been approved and a critical assessment of the existing data is necessary to balance benefits and harms and inform the development of evidence-based guidelines. METHODS: The American Gastroenterological Association formed a multidisciplinary group consisting of a Technical Review Panel and a Guideline Panel. The Technical Review Panel prioritized clinical questions and outcomes according to their importance for clinicians and patients and conducted an evidence review of systemic therapies in patients with advanced-stage HCC. The Grading of Recommendations Assessment, Development and Evaluation framework was used to assess evidence. The Guideline Panel reviewed the evidence and used the Evidence-to-Decision Framework to develop recommendations. RESULTS: The Panel reviewed the evidence, summarized in the Technical Review, for the following medications approved by the US Food and Drug Administration for HCC: first-line therapies: bevacizumab+atezolizumab, sorafenib, and lenvatinib; second-line therapies: cabozantinib, pembrolizumab, ramucirumab, and regorafenib; and other agents: bevacizumab, nivolumab, and nivolumab+ipilimumab. CONCLUSIONS: The Panel agreed on 11 recommendations focused on systemic therapy for HCC in patients who are not eligible for locoregional therapy or resection, those with metastatic disease and preserved liver function, those with poor liver function, and those on systemic therapy as adjuvant therapy.

Systematic review of research design and reporting of imaging studies applying convolutional neural networks for radiological cancer diagnosis.

OBJECTIVES: To perform a systematic review of design and reporting of imaging studies applying convolutional neural network models for radiological cancer diagnosis. METHODS: A comprehensive search of PUBMED, EMBASE, MEDLINE and SCOPUS was performed for published studies applying convolutional neural network models to radiological cancer diagnosis from January 1, 2016, to August 1, 2020. Two independent reviewers measured compliance with the Checklist for Artificial Intelligence in Medical Imaging (CLAIM). Compliance was defined as the proportion of applicable CLAIM items satisfied. RESULTS: One hundred eighty-six of 655 screened studies were included. Many studies did not meet the criteria for current design and reporting guidelines. Twenty-seven percent of studies documented eligibility criteria for their data (50/186, 95% CI 21-34%), 31% reported demographics for their study population (58/186, 95% CI 25-39%) and 49% of studies assessed model performance on test data partitions (91/186, 95% CI 42-57%). Median CLAIM compliance was 0.40 (IQR 0.33-0.49). Compliance correlated positively with publication year (ρ = 0.15, p = .04) and journal H-index (ρ = 0.27, p < .001). Clinical journals demonstrated higher mean compliance than technical journals (0.44 vs. 0.37, p < .001). CONCLUSIONS: Our findings highlight opportunities for improved design and reporting of convolutional neural network research for radiological cancer diagnosis. KEY POINTS: • Imaging studies applying convolutional neural networks (CNNs) for cancer diagnosis frequently omit key clinical information including eligibility criteria and population demographics. • Fewer than half of imaging studies assessed model performance on explicitly unobserved test data partitions. • Design and reporting standards have improved in CNN research for radiological cancer diagnosis, though many opportunities remain for further progress.

Clinical Outcomes of Acute Pancreatitis in Patients with Cirrhosis According to Liver Disease Severity Scores.

BACKGROUND: Literature on acute pancreatitis (AP) outcomes in patients with cirrhosis is limited. We aim to investigate the mortality and morbidity of AP in patients with cirrhosis. METHODS: We conducted a retrospective cohort study, and propensity score matching was done to match cirrhotic with non-cirrhotic patients on a 1:2 basis. Outcomes included inpatient mortality, organs failure, systemic inflammatory response syndrome, and length of hospital stay. We performed subgroup analysis of cirrhotics according to Child-Pugh and MELD scores. Multivariable logistic regression models were tested. RESULTS: From 819 AP patients, cirrhosis prevalence was 4.9% (40). There was no significant difference between cirrhotics and non-cirrhotics for inpatient mortality (7.5% vs. 1.3%, p = 0.1), severe AP (17.5% vs. 7.5%), shock (7.9% vs. 3%), respiratory failure (10% vs. 3.8%), need for intensive care unit (15% vs. 6.3%), systemic inflammatory response syndrome (SIRS) on admission (22.5% vs. 32.5%), and SIRS on day 2 (25% vs. 15%). Cirrhotics had similar rates of pancreatic necrosis, ileus, BISAP score, Marshall score, admission hematocrit, BUN, and hospital length of stay. Finally, cirrhotics who had severe AP, required ICU, and/or die in-hospital appeared to have more severe liver diseases (Child-C, higher MELD score > 17) and had lower AP severity scores (BISAP < 3, Marshall scores < 2). CONCLUSION: In our study, cirrhotics hospitalized with AP had similar morbidity and mortality when compared to non-cirrhotics.

Recent trends in the management of pelvic organ prolapse in Australia and New Zealand.

OBJECTIVE: To compare current practice in the management of female pelvic organ prolapse in Australia and New Zealand with that in 2007, and assess the impact on practice of the withdrawal of Prolift® and Prosima® mesh kits in 2015. MATERIALS AND METHODS: In early 2015, two invitations to participate in a survey, including a link to Surveymonkey, were emailed to 2506 Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) trainees and fellows. The online survey closely resembled a printed survey that was posted to RANZCOG trainees and fellows in 2007 and had additional questions relating to the impact of withdrawal of Prolift® and Prosima® products. RESULTS: Four-hundred-and-three doctors participated, giving a response rate of 16%. Native tissue repair was the procedure of choice for primary and recurrent prolapse of the anterior and posterior vaginal wall. An implant was used to treat 45% of anterior recurrences and 25% of posterior recurrences. Vaginal hysterectomy and repair were the procedures of choice for uterovaginal prolapse. Sacrospinous hysteropexy was the uterine preservation procedure of choice, preferred by 41%. For post-hysterectomy vault prolapse, sacrospinous colpopexy and vaginal repair was preferred by 65% of respondents. Between 2007 and 2015, there was a substantial decrease in respondents' usage of implants across all indications except for midurethral slings and sacrocolpo/hysteropexy. Forty-two percent of respondents changed their practice as a result of Prolift® and Prosima® being withdrawn. CONCLUSION: There is a trend toward increasing use of various native tissue prolapse repair procedures and midurethral slings, and less utilisation of transvaginal mesh for prolapse.

Diagnosis of Minimal Hepatic Encephalopathy Using Stroop EncephalApp: A Multicenter US-Based, Norm-Based Study.

OBJECTIVES: Diagnosing minimal hepatic encephalopathy (MHE) is challenging, and point-of-care tests are needed. Stroop EncephalApp has been validated for MHE diagnosis in single-center studies. The objective of the study was to validate EncephalApp for MHE diagnosis in a multicenter study. METHODS: Outpatient cirrhotics (with/without prior overt hepatic encephalopathy (OHE)) and controls from three sites (Virginia (VA), Ohio (OH), and Arkansas (AR)) underwent EncephalApp and two gold standards, psychometric hepatic encephalopathy score (PHES) and inhibitory control test (ICT). Age-/gender-/education-adjusted values for EncephalApp based on direct norms, and based on ICT and PHES, were defined. Patients were followed, and EncephalApp cutoff points were used to determine OHE prediction. These cutoff points were then used in a separate VA-based validation cohort. RESULTS: A total of 437 cirrhotics (230 VA, 107 OH, 100 AR, 36% OHE, model for end-stage liver disease (MELD) score 11) and 308 controls (103 VA, 100 OH, 105 AR) were included. Using adjusted variables, MHE was present using EncephalApp based on norms in 51%, EncephalApp based on PHES in 37% (sensitivity 80%), and EncephalApp based on ICT in 54% of patients (sensitivity 70%). There was modest/good agreement between sites on EncephalApp MHE diagnosis using the three methods. OHE developed in 13% of patients, which was predicted by EncephalApp independent of the MELD score. In the validation cohort of 121 VA cirrhotics, EncephalApp directly and based on gold standards remained consistent for MHE diagnosis with >70% sensitivity. CONCLUSIONS: In this multicenter study, EncephalApp, using adjusted population norms or in the context of existing gold standard tests, had good sensitivity for MHE diagnosis and predictive capability for OHE development.

We make predictions about eye of origin of visual input: Visual mismatch negativity from binocular rivalry.

The visual mismatch negativity (vMMN) is a negative component of event-related potentials (ERPs). It occurs when an infrequent visual stimulus, a deviant, is randomly and unpredictably presented in a sequence of frequent visual stimuli, the standards, and is thought to reflect prediction and prediction error of visual input. We investigated the sensitivity of vMMN to eye of origin (utrocular) information as well as to orientation information. We presented 80% of binocular rivalry standards (one grating to one eye and an identical, orthogonally oriented grating to the other eye), and 20% of deviants, either by swapping the gratings between the eyes to change the eye of origin of the gratings (an eye-swap deviant) or by rotating the gratings by 45° to change the orientation of the gratings (an orientation deviant). We found an orientation vMMN that was maximal at about 250 ms and an eye-swap vMMN that was maximal at about 380 ms. We also found deviance-related activity to both sorts of stimuli earlier than is traditionally defined as a vMMN. We used standardized low-resolution brain electromagnetic tomography (sLORETA) to localize each vMMN component and found similar sources for both vMMNs in occipital and frontal areas of the brain but differences in parietal and temporal areas. We conclude that eye of origin information can be used to elicit vMMN, that eye-swap vMMN is different to orientation vMMN, and that vMMN can be generated from information of which observers are unaware.

Long-term outcomes of laparoscopic repair of cystocoele.

BACKGROUND: There is little information on the effectiveness of laparoscopic techniques for native tissue repair of cystocoele. AIM: To assess the long-term outcome of laparoscopic cystocoele repair. METHODS: Two hundred and twenty-three women with symptomatic pelvic organ prolapse underwent laparoscopic paravaginal repair and treatment of associated conditions. Women were assessed pre-operatively and postoperatively at 6 weeks, 6 months, 12 months and then annually or biannually with pelvic organ prolapse quantification (POPQ) and subjective assessment at each visit. RESULTS: Median follow-up was 5.2 years (range: 1-12 years) with 140 women (63%) followed for at least 5 years. During follow-up, 79% of women developed prolapse of at least POPQ stage 2 in one or more compartments and 58% became symptomatic again. Overall, 48% underwent further prolapse surgery, but only 24% of women had an anterior prolapse beyond the hymen. Thirty per cent eventually had a further cystocoele repair. CONCLUSION: Long-term follow-up of laparoscopic cystocoele repair shows that cystocoeles are difficult to repair successfully.

Multicentre pilot randomised control trial of a self-directed exergaming intervention for poststroke upper limb rehabilitation: research protocol.

INTRODUCTION: Technology-facilitated, self-directed upper limb (UL) rehabilitation, as an adjunct to conventional care, could enhance poststroke UL recovery compared with conventional care alone, without imposing additional resource burden. The proposed pilot randomised controlled trial (RCT) aims to assess whether stroke survivors will engage in self-directed UL training, explore factors associated with intervention adherence and evaluate the study design for an RCT testing the efficacy of a self-directed exer-gaming intervention for UL recovery after stroke. METHODS AND ANALYSIS: This is a multicentre, internal pilot RCT; parallel design, with nested qualitative methods. The sample will consist of stroke survivors with UL paresis, presenting within the previous 30 days. Participants randomised to the intervention group will be trained to use an exergaming device and will be supported to adopt this as part of their self-directed rehabilitation (ie, without formal support/supervision) for a 3-month period. The primary outcome will be the Fugl Meyer Upper Extremity Assessment (FM-UE) at 6 months poststroke. Secondary outcomes are the Action Research Arm Test (ARAT), the Barthel Index and the Modified Rankin Scale. Assessment time points will be prior to randomisation (0-1 month poststroke), 3 months and 6 months poststroke. A power calculation to inform sample size required for a definitive RCT will be conducted using FM-UE data from the sample across 0-6 months time points. Semistructured qualitative interviews will examine factors associated with intervention adoption. Reflexive thematic analysis will be used to code qualitative interview data and generate key themes associated with intervention adoption. ETHICS AND DISSEMINATION: The study protocol (V.1.9) was granted ethical approval by the Health Research Authority, Health and Care Research Wales, and the London- Harrow Research Ethics Committee (ref. 21/LO/0054) on 19 May 2021. Trial results will be submitted for publication in peer-reviewed journals, presented at national and international stroke meetings and conferences and disseminated among stakeholder communities. TRIAL REGISTRATION NUMBER: NCT04475692.

Effect of uterine preservation on outcome of laparoscopic uterosacral suspension.

STUDY OBJECTIVE: To compare the objective outcome of laparoscopic uterosacral hysteropexy with that of hysterectomy combined with laparoscopic uterosacral colpopexy. DESIGN: Retrospective cohort study, 1999-2010 (Canadian Task Force classification II-2). SETTING: University hospital in South Australia. PATIENTS: Women with uterovaginal prolapse who had undergone laparoscopic uterosacral hysteropexy (n = 104) or laparovaginal hysterectomy with uterosacral colpopexy (n = 160). Apical suspension procedures were subdivided into prophylactic (Pelvic Organ Prolapse Quantification System [POP-Q] stage 1 apical descent, with stage ≥2 prolapse in an adjacent compartment) and therapeutic (POP-Q stage ≥2 apical descent, with or without adjacent compartment prolapse). INTERVENTIONS: All patients were assessed via POP-Q scoring preoperatively and postoperatively at 6 weeks, 6 months, annually, and then biannually. Recurrence of bulge symptoms and need for repeat treatment were recorded. MEASUREMENTS AND MAIN RESULTS: Demographic data, preoperative degree of prolapse, and percentages of prophylactic and therapeutic procedures were similar in both groups. With a median follow-up of 2.5 years, objective success rates (POP-Q stage <2 in all compartments) for uterosacral hysteropexy were 53% for prophylactic procedures and 41% for therapeutic procedures, and for hysterectomy with uterosacral colpopexy were 66% for prophylactic procedures and 59% for therapeutic procedures. Repeat operation rates overall were 28% for hysteropexy and 21% for hysterectomy with colpopexy. Failures at the apex specifically were 27% for hysteropexy and 11% for hysterectomy with colpopexy (p < .02). CONCLUSION: Hysterectomy with laparoscopic uterosacral colpopexy produced better objective success rates than did laparoscopic uterosacral hysteropexy; however, repeat operation rates were not significantly different.

Examining Usability, Acceptability, and Adoption of a Self-Directed, Technology-Based Intervention for Upper Limb Rehabilitation After Stroke: Cohort Study.

BACKGROUND: Upper limb (UL) recovery after stroke is strongly dependent upon rehabilitation dose. Rehabilitation technologies present pragmatic solutions to dose enhancement, complementing therapeutic activity within conventional rehabilitation, connecting clinicians with patients remotely, and empowering patients to drive their own recovery. To date, rehabilitation technologies have been poorly adopted. Understanding the barriers to adoption may shape strategies to enhance technology use and therefore increase rehabilitation dose, thus optimizing recovery potential. OBJECTIVE: We examined the usability, acceptability, and adoption of a self-directed, exercise-gaming technology within a heterogeneous stroke survivor cohort and investigated how stroke survivor characteristics, technology usability, and attitudes toward technology influenced adoption. METHODS: A feasibility study of a novel exercise-gaming technology for self-directed UL rehabilitation in early subacute stroke survivors (N=30) was conducted in an inpatient, acute hospital setting. Demographic and clinical characteristics were recorded; participants' performance in using the system (usability) was assessed using a 4-point performance rating scale (adapted from the Barthel index), and adherence with the system was electronically logged throughout the trial. The technology acceptance model was used to formulate a survey examining the acceptability of the system. Spearman rank correlations were used to examine associations between participant characteristics, user performance (usability), end-point technology acceptance, and intervention adherence (adoption). RESULTS: The technology was usable for 87% (n=26) of participants, and the overall technology acceptance rating was 68% (95% CI 56%-79%). Participants trained with the device for a median of 26 (IQR 16-31) minutes daily over an enrollment period of 8 (IQR 5-14) days. Technology adoption positively correlated with user performance (usability) (ρ=0.55; 95% CI 0.23-0.75; P=.007) and acceptability as well as domains of perceived usefulness (ρ=0.42; 95% CI 0.09-0.68; P=.03) and perceived ease of use (ρ=0.46; 95% CI 0.10-0.74; P=.02). Technology acceptance decreased with increased global stroke severity (ρ=-0.56; 95% CI -0.79 to -0.22; P=.007). CONCLUSIONS: This technology was usable and acceptable for the majority of the cohort, who achieved an intervention dose with technology-facilitated, self-directed UL training that exceeded conventional care norms. Technology usability and acceptability were determinants of adoption and appear to be mediated by stroke severity. The results demonstrate the importance of selecting technologies for stroke survivors on the basis of individual needs and abilities, as well as optimizing the accessibility of technologies for the target user group. Facilitating changes in stroke survivors' beliefs and attitudes toward rehabilitation technologies may enhance adoption. Further work is needed to understand how technology can be optimized to benefit those with more severe stroke.

Weakly supervised segmentation models as explainable radiological classifiers for lung tumour detection on CT images.

PURPOSE: Interpretability is essential for reliable convolutional neural network (CNN) image classifiers in radiological applications. We describe a weakly supervised segmentation model that learns to delineate the target object, trained with only image-level labels ("image contains object" or "image does not contain object"), presenting a different approach towards explainable object detectors for radiological imaging tasks. METHODS: A weakly supervised Unet architecture (WSUnet) was trained to learn lung tumour segmentation from image-level labelled data. WSUnet generates voxel probability maps with a Unet and then constructs an image-level prediction by global max-pooling, thereby facilitating image-level training. WSUnet's voxel-level predictions were compared to traditional model interpretation techniques (class activation mapping, integrated gradients and occlusion sensitivity) in CT data from three institutions (training/validation: n = 412; testing: n = 142). Methods were compared using voxel-level discrimination metrics and clinical value was assessed with a clinician preference survey on data from external institutions. RESULTS: Despite the absence of voxel-level labels in training, WSUnet's voxel-level predictions localised tumours precisely in both validation (precision: 0.77, 95% CI: [0.76-0.80]; dice: 0.43, 95% CI: [0.39-0.46]), and external testing (precision: 0.78, 95% CI: [0.76-0.81]; dice: 0.33, 95% CI: [0.32-0.35]). WSUnet's voxel-level discrimination outperformed the best comparator in validation (area under precision recall curve (AUPR): 0.55, 95% CI: [0.49-0.56] vs. 0.23, 95% CI: [0.21-0.25]) and testing (AUPR: 0.40, 95% CI: [0.38-0.41] vs. 0.36, 95% CI: [0.34-0.37]). Clinicians preferred WSUnet predictions in most instances (clinician preference rate: 0.72 95% CI: [0.68-0.77]). CONCLUSION: Weakly supervised segmentation is a viable approach by which explainable object detection models may be developed for medical imaging. CRITICAL RELEVANCE STATEMENT: WSUnet learns to segment images at voxel level, training only with image-level labels. A Unet backbone first generates a voxel-level probability map and then extracts the maximum voxel prediction as the image-level prediction. Thus, training uses only image-level annotations, reducing human workload. WSUnet's voxel-level predictions provide a causally verifiable explanation for its image-level prediction, improving interpretability. KEY POINTS: • Explainability and interpretability are essential for reliable medical image classifiers. • This study applies weakly supervised segmentation to generate explainable image classifiers. • The weakly supervised Unet inherently explains its image-level predictions at voxel level.

Laparoscopic supralevator repair for combined apical and posterior compartment prolapse.

STUDY OBJECTIVE: To analyze the objective outcome of laparoscopic supralevator repair in the treatment of rectoenterocele with the Pelvic Organ Prolapse Quantification (POPQ) system. STUDY DESIGN: Retrospective cohort study 1999-2009 (Canadian Taskforce Classification II-2). SETTING: University hospital in South Australia. PATIENTS: A total of 166 women with a median age of 63 years (range 36-89) who underwent laparoscopic supralevator repair for rectoenterocele and treatment of associated conditions over a 10-year period. INTERVENTIONS: All patients were assessed with the POPQ scoring system before surgery and at 6 weeks, 6 months, annually, and biannually after surgery. MEASUREMENTS AND MAIN RESULTS: The median operating time was 151 minutes (range 35-390); median blood loss was 50 mL (range 50-600); and median hospital stay was 4 days (range 1-14). Four women, 2 of whom required laparotomy, had a major complication. Ten women (6%) needed day surgery to treat vaginal granulations or suture exposure. With a median follow-up time of 45 months (interquartile range 16-67) the overall objective success rate was 63% according to National Institute of Health criteria. The median time to failure was 24 months. Of 61 objective failures, 23 required further prolapse surgery, representing a 14% reoperation rate. CONCLUSION: Laparoscopic supralevator repair is a safe and effective procedure for the treatment of rectoenterocele.

Radiomic assessment of oesophageal adenocarcinoma: a critical review of 18F-FDG PET/CT, PET/MRI and CT.

OBJECTIVES: Radiomic models present an avenue to improve oesophageal adenocarcinoma assessment through quantitative medical image analysis. However, model selection is complicated by the abundance of available predictors and the uncertainty of their relevance and reproducibility. This analysis reviews recent research to facilitate precedent-based model selection for prospective validation studies. METHODS: This analysis reviews research on 18F-FDG PET/CT, PET/MRI and CT radiomics in oesophageal adenocarcinoma between 2016 and 2021. Model design, testing and reporting are evaluated according to the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) score and Radiomics Quality Score (RQS). Key results and limitations are analysed to identify opportunities for future research in the area. RESULTS: Radiomic models of stage and therapeutic response demonstrated discriminative capacity, though clinical applications require greater sensitivity. Although radiomic models predict survival within institutions, generalisability is limited. Few radiomic features have been recommended independently by multiple studies. CONCLUSIONS: Future research must prioritise prospective validation of previously proposed models to further clinical translation.

Sparse Regression in Cancer Genomics: Comparing Variable Selection and Predictions in Real World Data.

BACKGROUND: Evaluation of gene interaction models in cancer genomics is challenging, as the true distribution is uncertain. Previous analyses have benchmarked models using synthetic data or databases of experimentally verified interactions - approaches which are susceptible to misrepresentation and incompleteness, respectively. The objectives of this analysis are to (1) provide a real-world data-driven approach for comparing performance of genomic model inference algorithms, (2) compare the performance of LASSO, elastic net, best-subset selection, L 0 L 1 penalisation and L 0 L 2 penalisation in real genomic data and (3) compare algorithmic preselection according to performance in our benchmark datasets to algorithmic selection by internal cross-validation. METHODS: Five large ( n 4000 ) genomic datasets were extracted from Gene Expression Omnibus. 'Gold-standard' regression models were trained on subspaces of these datasets ( n 4000 , p = 500 ). Penalised regression models were trained on small samples from these subspaces ( n ∈ { 25 , 75 , 150 } , p = 500 ) and validated against the gold-standard models. Variable selection performance and out-of-sample prediction were assessed. Penalty 'preselection' according to test performance in the other 4 datasets was compared to selection internal cross-validation error minimisation. RESULTS: L 1 L 2 -penalisation achieved the highest cosine similarity between estimated coefficients and those of gold-standard models. L 0 L 2 -penalised models explained the greatest proportion of variance in test responses, though performance was unreliable in low signal:noise conditions. L 0 L 2 also attained the highest overall median variable selection F1 score. Penalty preselection significantly outperformed selection by internal cross-validation in each of 3 examined metrics. CONCLUSIONS: This analysis explores a novel approach for comparisons of model selection approaches in real genomic data from 5 cancers. Our benchmarking datasets have been made publicly available for use in future research. Our findings support the use of L 0 L 2 penalisation for structural selection and L 1 L 2 penalisation for coefficient recovery in genomic data. Evaluation of learning algorithms according to observed test performance in external genomic datasets yields valuable insights into actual test performance, providing a data-driven complement to internal cross-validation in genomic regression tasks.

Seeing polarization of light with the naked eye.

Many readers may know that scores of animal species sense the polarization of light for purposes including navigation, predation, and communication1. It is commonly thought that humans lack any sensitivity to polarization of light (e.g., Morehouse2). We hope to convince you otherwise by describing three examples where humans can detect polarization of light with the naked eye, by showing you how to see it yourself, and by describing its uses.