Artificial intelligence for MRI stroke detection: a systematic review and meta-analysis.

OBJECTIVES: This systematic review and meta-analysis aimed to assess the stroke detection performance of artificial intelligence (AI) in magnetic resonance imaging (MRI), and additionally to identify reporting insufficiencies. METHODS: PRISMA guidelines were followed. MEDLINE, Embase, Cochrane Central, and IEEE Xplore were searched for studies utilising MRI and AI for stroke detection. The protocol was prospectively registered with PROSPERO (CRD42021289748). Sensitivity, specificity, accuracy, and area under the receiver operating characteristic (ROC) curve were the primary outcomes. Only studies using MRI in adults were included. The intervention was AI for stroke detection with ischaemic and haemorrhagic stroke in separate categories. Any manual labelling was used as a comparator. A modified QUADAS-2 tool was used for bias assessment. The minimum information about clinical artificial intelligence modelling (MI-CLAIM) checklist was used to assess reporting insufficiencies. Meta-analyses were performed for sensitivity, specificity, and hierarchical summary ROC (HSROC) on low risk of bias studies. RESULTS: Thirty-three studies were eligible for inclusion. Fifteen studies had a low risk of bias. Low-risk studies were better for reporting MI-CLAIM items. Only one study examined a CE-approved AI algorithm. Forest plots revealed detection sensitivity and specificity of 93% and 93% with identical performance in the HSROC analysis and positive and negative likelihood ratios of 12.6 and 0.079. CONCLUSION: Current AI technology can detect ischaemic stroke in MRI. There is a need for further validation of haemorrhagic detection. The clinical usability of AI stroke detection in MRI is yet to be investigated. CRITICAL RELEVANCE STATEMENT: This first meta-analysis concludes that AI, utilising diffusion-weighted MRI sequences, can accurately aid the detection of ischaemic brain lesions and its clinical utility is ready to be uncovered in clinical trials. KEY POINTS: There is a growing interest in AI solutions for detection aid. The performance is unknown for MRI stroke assessment. AI detection sensitivity and specificity were 93% and 93% for ischaemic lesions. There is limited evidence for the detection of patients with haemorrhagic lesions. AI can accurately detect patients with ischaemic stroke in MRI.

Breast cancer detection accuracy of AI in an entire screening population: a retrospective, multicentre study.

BACKGROUND: Artificial intelligence (AI) systems are proposed as a replacement of the first reader in double reading within mammography screening. We aimed to assess cancer detection accuracy of an AI system in a Danish screening population. METHODS: We retrieved a consecutive screening cohort from the Region of Southern Denmark including all participating women between Aug 4, 2014, and August 15, 2018. Screening mammograms were processed by a commercial AI system and detection accuracy was evaluated in two scenarios, Standalone AI and AI-integrated screening replacing first reader, with first reader and double reading with arbitration (combined reading) as comparators, respectively. Two AI-score cut-off points were applied by matching at mean first reader sensitivity (AIsens) and specificity (AIspec). Reference standard was histopathology-proven breast cancer or cancer-free follow-up within 24 months. Coprimary endpoints were sensitivity and specificity, and secondary endpoints were positive predictive value (PPV), negative predictive value (NPV), recall rate, and arbitration rate. Accuracy estimates were calculated using McNemar's test or exact binomial test. RESULTS: Out of 272,008 screening mammograms from 158,732 women, 257,671 (94.7%) with adequate image data were included in the final analyses. Sensitivity and specificity were 63.7% (95% CI 61.6%-65.8%) and 97.8% (97.7-97.8%) for first reader, and 73.9% (72.0-75.8%) and 97.9% (97.9-98.0%) for combined reading, respectively. Standalone AIsens showed a lower specificity (-1.3%) and PPV (-6.1%), and a higher recall rate (+ 1.3%) compared to first reader (p < 0.0001 for all), while Standalone AIspec had a lower sensitivity (-5.1%; p < 0.0001), PPV (-1.3%; p = 0.01) and NPV (-0.04%; p = 0.0002). Compared to combined reading, Integrated AIsens achieved higher sensitivity (+ 2.3%; p = 0.0004), but lower specificity (-0.6%) and PPV (-3.9%) as well as higher recall rate (+ 0.6%) and arbitration rate (+ 2.2%; p < 0.0001 for all). Integrated AIspec showed no significant difference in any outcome measures apart from a slightly higher arbitration rate (p < 0.0001). Subgroup analyses showed higher detection of interval cancers by Standalone AI and Integrated AI at both thresholds (p < 0.0001 for all) with a varying composition of detected cancers across multiple subgroups of tumour characteristics. CONCLUSIONS: Replacing first reader in double reading with an AI could be feasible but choosing an appropriate AI threshold is crucial to maintaining cancer detection accuracy and workload.

Population-wide evaluation of artificial intelligence and radiologist assessment of screening mammograms.

OBJECTIVES: To validate an AI system for standalone breast cancer detection on an entire screening population in comparison to first-reading breast radiologists. MATERIALS AND METHODS: All mammography screenings performed between August 4, 2014, and August 15, 2018, in the Region of Southern Denmark with follow-up within 24 months were eligible. Screenings were assessed as normal or abnormal by breast radiologists through double reading with arbitration. For an AI decision of normal or abnormal, two AI-score cut-off points were applied by matching at mean sensitivity (AIsens) and specificity (AIspec) of first readers. Accuracy measures were sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and recall rate (RR). RESULTS: The sample included 249,402 screenings (149,495 women) and 2033 breast cancers (72.6% screen-detected cancers, 27.4% interval cancers). AIsens had lower specificity (97.5% vs 97.7%; p < 0.0001) and PPV (17.5% vs 18.7%; p = 0.01) and a higher RR (3.0% vs 2.8%; p < 0.0001) than first readers. AIspec was comparable to first readers in terms of all accuracy measures. Both AIsens and AIspec detected significantly fewer screen-detected cancers (1166 (AIsens), 1156 (AIspec) vs 1252; p < 0.0001) but found more interval cancers compared to first readers (126 (AIsens), 117 (AIspec) vs 39; p < 0.0001) with varying types of cancers detected across multiple subgroups. CONCLUSION: Standalone AI can detect breast cancer at an accuracy level equivalent to the standard of first readers when the AI threshold point was matched at first reader specificity. However, AI and first readers detected a different composition of cancers. CLINICAL RELEVANCE STATEMENT: Replacing first readers with AI with an appropriate cut-off score could be feasible. AI-detected cancers not detected by radiologists suggest a potential increase in the number of cancers detected if AI is implemented to support double reading within screening, although the clinicopathological characteristics of detected cancers would not change significantly. KEY POINTS: • Standalone AI cancer detection was compared to first readers in a double-read mammography screening population. • Standalone AI matched at first reader specificity showed no statistically significant difference in overall accuracy but detected different cancers. • With an appropriate threshold, AI-integrated screening can increase the number of detected cancers with similar clinicopathological characteristics.

Colon capsule retention in a patient with large bowel stenosis due to diverticulosis - a case report.

Capsule retention is a rare complication to capsule endoscopy. It is often asymptomatic and resolves itself spontaneously. In some cases, endoscopy or surgery is necessary to remove the capsule. Cases of capsule retention in the colon are not reported very often and precautions in subsequent investigations must be considered. We present a case of a 74-year-old woman who underwent colon capsule endoscopy (CCE) due to a positive fecal immunochemical test. The CCE was incomplete, and the patient was referred to conventional colonoscopy, which was complicated by perforation of the large bowel during the procedure. This lead to a CT scan showing the capsule proximal to a stenotic colon segment which was subsequently surgically removed. In patients who report lack of capsule excretion and stenosis is not verified, colonoscopy can be carried out, but with caution.

Renal functional outcomes after robot-assisted partial nephrectomy and percutaneous cryoablation of clinical T1 renal cell carcinoma - A prospective study.

Objectives: The objective of this study was to investigate renal function after robot-assisted partial nephrectomy (RAPN) and percutaneous cryoablation (PCA) in clinical stage T1 (cT1) renal cell carcinoma (RCC) and evaluate the relationship between baseline renal function and renal functional outcome. Material and Methods: Patients with cT1 RCC treated with RAPN or PCA were prospectively enrolled between June 2019 and January 2021. Renal function was evaluated using estimated glomerular filtration rate, Tc-99m diethylenetriamine-pentaacetate plasma clearance, Tc-99m mercaptoacetyltriglycine renography, and renal volume at baseline and 6 months after treatment. Results: Fifty-six patients were included (18 RAPN, 38 PCA). PCA patients had a significantly higher age (68.5 years; P = 0.019) and Charlson comorbidity index (3.0; P = 0.007). Tumor characteristics did not differ significantly between RAPN and PCA. Total renal volume decreased significantly after PCA (-18.2 cm3; P = 0.001). Baseline chronic disease stage IIIb-IV leads to a greater reduction in renal volume (-31.8 cm3; P = 0.003) but not other renal function measures. Renal function declined significantly after either treatment without significant differences between RAPN and PCA. Conclusion: This study found a small, similar decrease in renal function 6 months after RAPN or PCA, despite significant differences in baseline patient characteristics. Reduced renal function at baseline did not lead to a worse renal functional outcome.

[Amphetamine-induced toxic encephalopathy].

In this case report, a 40-year-old woman presented with altered mentation, central facial palsy and hemiparesis. Acute CT scan showed several hypodense, rounded areas in both hemispheres. Urine toxicology tested positive for amphetamine. Subsequent cerebral MRI had hyperintense T2-weighted fluid-attenuated inverse recovery (FLAIR) lesions in both hemispheres, indicating demyelinating disease. A biopsy was made from one of the lesions. The final diagnosis was toxic encephalopathy.

[Artificial intelligence for cancer detection in breast cancer screening].

Artificial intelligence (AI) has the potential to increase quality and efficiency of breast cancer screening. Recent studies have provided comparative data on AI versus human performance in cancer detection with encouraging results, and commercially available AI systems are used worldwide as a clinical tool for mammography screening. There are, however, a number of methodological concerns in relation to the evaluation of AI systems. This review discusses these aspects as well as the opportunities and challenges of clinical validation and implementation in breast cancer screening practice.

Percutaneous cryoablation of a solitary, soft-tissue metastasis from renal cell carcinoma: a new local minimal invasive curative treatment.

Cryoablation is a well-established treatment option, proven to be successful in treating local renal cell carcinoma (RCC). We treated a 67-year-old man in an outpatient setting with late onset of a 25 mm solitary soft-tissue metastasis of from RCC with cryoablation. The treatment was performed under sedation and in local anaesthesia. There were no complications during the procedure. The patient did not experience any adverse effects to the treatment. He was able to resume his normal daily routines the day after his treatment. A follow-up CT scan at 3, 8 and 12 months after treatment reported sufficient cryoablation and no sign of recurrence or other metastases.

Body Mass Index, Intensive Blood Pressure Management, and Cardiovascular Events in the SPRINT Trial.

BACKGROUND: It is unclear whether intensive blood pressure management is well-tolerated and affects risk uniformly across the body mass index (BMI) spectrum. METHODS: The randomized, controlled Systolic Blood Pressure Intervention Trial (SPRINT) included 9361 individuals ≥50 years of age at high cardiovascular risk, without diabetes mellitus, with systolic blood pressure between 130 and 180 mmHg. Participants were randomized to intensive vs standard antihypertensive treatment and evaluated for the primary composite efficacy endpoint of acute coronary syndromes, stroke, heart failure, or cardiovascular death. The primary safety endpoint was serious adverse events. We used restricted cubic splines to determine the relationship between BMI, response to intensive blood pressure lowering, and clinical outcomes in SPRINT. RESULTS: Body mass index could be calculated for 9284 (99.2%) individuals. Mean BMI was similar between the 2 treatment groups (intensive group 29.9±5.8 kg/m2 vs standard group 29.8± 5.7 kg/m2; P = 0.39). Median follow-up was 3.3 years (range 0-4.8 years). Body mass index had a significant, J-shaped association with risk of all-cause mortality, stroke, and serious adverse events (P < .05 for all), but these were no longer significant after accounting for key clinical factors (P > .05 for all). Intensive blood pressure lowering reduced the primary efficacy endpoint and increased the primary safety endpoint compared with standard targets, consistently across the BMI spectrum (Pinteraction > .05). CONCLUSION: The overall efficacy and safety of intensive blood pressure lowering did not appear to be modified by baseline BMI among high-risk older adults.

Validation of three-dimensional wound measurements using a novel 3D-WAM camera.

To monitor wound healing, it is essential to obtain accurate and reliable wound measurements. Various methods have been used to measure wound size including three-dimensional (3D) measurement devices enabling wound assessment from a volume perspective. However, the currently available methods are inaccurate, costly, or complicated to use. As a consequence, we have developed a 3D-wound assessment monitor (WAM) camera, which is able to measure wound size in three-dimension and to assess wound characteristics. The aim of the study was to assess the intrarater and interrater reliability of the 3D wound measurements using the 3D camera and to compare these with traditional measurement methods. Four raters measured 48 wounds using the 3D camera, digital imaging method (2D area), and gel injection into the wound cavity (volume). The data were analyzed using linear mixed effect model. Intraclass and interclass correlation coefficient (ICC) and Bland-Altman plots were used to assess intrarater and interrater reliability for the 3D camera and agreement between the methods. The Bland-Altman plots for intrarater reliability showed minor differences between the measurements, especially the 3D area and perimeter measurements. Moreover, ICCs were very high for both the intrarater and interrater reliability for the 2D area, 3D area, and perimeter measurements (ICCs > 0.99), although slightly lower for the volume measurements (ICC = 0.946-0.950). Finally, a high agreement was found between the 3D camera and the traditional methods (2D area and volume) assessed by narrow 95% prediction intervals and high ICCs above 0.97. In conclusion, the 3D-WAM camera is an accurate and reliable method, which is useful for several types of wounds. However, the volume measurements were primarily useful in large, deep wounds. Moreover, the 3D images are based on digital technology and therefore carry the possibility for use in remote settings.

Cost-effectiveness of telemonitoring of diabetic foot ulcer patients.

This study compared the cost-effectiveness of telemonitoring with standard monitoring for patients with diabetic foot ulcers. The economic evaluation was nested within a pragmatic randomised controlled trial. A total of 374 patients were randomised to either telemonitoring or standard monitoring. Telemonitoring consisted of two tele-consultations in the patient's own home and one consultation at the outpatient clinic; standard monitoring consisted of three outpatient clinic consultations. Total healthcare costs were estimated over a 6-month period at individual patient level, from a healthcare sector perspective. The bootstrap method was used to calculate the incremental cost-effectiveness ratio, and one-way sensitivity analyses were performed. Telemonitoring costs were found to be €2039 less per patient compared to standard monitoring; however, this difference was not statistically significant. Amputation rate was similar in the two groups. In conclusion, a telemonitoring service in this form had similar costs and effects as standard monitoring.

Outbreak of Pseudomonas aeruginosa bacteraemia in a haematology department.

INTRODUCTION: Infection by Pseudomonas aeruginosa represents a major cause of morbidity and mortality among immunocompromised patients. In Denmark, an increase in P. aeruginosa isolates from blood cultures from a haematology department prompted a hygienic audit in 2007. METHODS: Blood cultures that tested positive for P. aeruginosa were collected from the laboratory information system (MADS, Skejby Hospital, Aarhus, Denmark). Environmental samples were obtained from shower heads in the department. The genotype was established by pulse field gel electrophoresis (PFGE). An audit was conducted during the outbreak and 12 months later. The audits were conducted by the method of direct observation. RESULTS: Several PFGE types were involved with no clear association to isolates from environmental samples. The audit revealed poor hygiene related to the handling of central venous catheters. After optimising catheter hygiene, the number of P. aeruginosa bacteraemia cases fell significantly. CONCLUSION: Since no clear association between patient and environmental genotype was established, it was suspected that central venous catheters were the main portal of entry. This was further supported by a simultaneous decline in bacteraemia cases with coagulase-negative staphylococci. Though several hygienic precautions were taken, the increased focus on disinfection of hubs and injection ports was presumably the more important element. FUNDING: not relevant. TRIAL REGISTRATION: not relevant.

eHealth in Wound Care: from conception to implementation.

Executive summary Purpose This document aims to provide wound care clinicians with a rapid and structured overview of the key issues related to use of eHealth applications (telemedicine and telehealth) within wound care. This includes: An overview of terminology and available literature Guidance on the methodology for evaluation of eHealth solutions An introduction to and discussion of the potential benefits of eHealth technologies in wound care, and the possible barriers to their implementation Recommendations for ensuring a good implementation process and supporting involvement of wound care professionals in safeguarding that eHealth solutions meet the needs of the patients. Methodology The document sections lean on the structure and focus areas of the Model for ASsessment of Telemedicine (MAST) which defines crucial items to evaluate an eHealth application. The content of the document is developed on the basis of a literature review, identifying available documentation for use of eHealth solutions in wound care. Furthermore, it draws on various key documents recently published on the general development, evaluation and implementation of eHealth solutions. These include valuable up-to-date information relevant for any group of clinicians wishing to follow and influence the way eHealth solutions are integrated into clinical practice. Findings and conclusions The literature review revealed that the amount and level of evidence for use of eHealth applications in wound care is still limited. Some MAST domains are not examined in any of the available studies. Thus, more research is required to identify the potential benefits and harms to patients, and the possible challenges related to implementation of eHealth solutions in wound care. Potential barriers and facilitators for the implementation of eHealth applications into wound care practice are identified in the document, and these may all either enhance or impede the process. However, the available research does demonstrate patient satisfaction, improved access to health services for all client cohorts, and increased job satisfaction for clinicians. The document recommends that wound management clinicians, considering the use of eHealth applications in their clinical practice, consult widely and conduct regular evaluation of the outcomes to ensure efficient implementation of these services. To support this approach, steps to ensure a good implementation process within a given organisation have been proposed. These are synthesised into a three circle model.