Hierarchical segmentation of surgical scenes in laparoscopy.

PURPOSE: Segmentation of surgical scenes may provide valuable information for real-time guidance and post-operative analysis. However, in some surgical video frames there is unavoidable ambiguity, leading to incorrect predictions of class or missed detections. In this work, we propose a novel method that alleviates this problem by introducing a hierarchy and associated hierarchical inference scheme that allows broad anatomical structures to be predicted when fine-grained structures cannot be reliably distinguished. METHODS: First, we formulate a multi-label segmentation loss informed by a hierarchy of anatomical classes and then train a network using this. Subsequently, we use a novel leaf-to-root inference scheme ("Hiera-Mix") to determine the trade-off between label confidence and granularity. This method can be applied to any segmentation model. We evaluate our method using a large laparoscopic cholecystectomy dataset with 65,000 labelled frames. RESULTS: We observed an increase in per-structure detection F1 score for the critical structures, when evaluated across their sub-hierarchies, compared to the baseline method: 6.0% for the cystic artery and 2.9% for the cystic duct, driven primarily by increases in precision of 11.3% and 4.7%, respectively. This corresponded to visibly improved segmentation outputs, with better characterisation of the undissected area containing the critical structures and fewer inter-class confusions. For other anatomical classes, which did not stand to benefit from the hierarchy, performance was unimpaired. CONCLUSION: Our proposed hierarchical approach improves surgical scene segmentation in frames with ambiguity, by more suitably reflecting the model's parsing of the scene. This may be beneficial in applications of surgical scene segmentation, including recent advancements towards computer-assisted intra-operative guidance.

Paraspinal soft tissue edema ratio: An accurate marker for early lumbar spine spondylodiscitis on an unenhanced MRI.

PURPOSE: MRI is currently the gold standard imaging modality in the diagnosis of lumbar spine discitis/osteomyelitis. However, even with supportive clinical and laboratory data, the accuracy of MRI remains limited by several degenerative and inflammatory mimics, such that it continues to represent a challenge for radiologists. This study reports a new quantitative imaging marker of lumbar paraspinal soft tissue edema which shows significant accuracy for spondylodiscitis. METHODS: Thirty-five patients with equivocal MRI findings of lumbar discitis/osteomyelitis vs endplate degenerative changes were reviewed over a 24-month period. Patients with a history of surgery, fractures/recent trauma, signs of advanced infection such as abscesses, phlegmon or severe osseous destruction were excluded. Two ABR board certified neuroradiologists who were blinded to the final diagnosis evaluated a new marker; the superior-inferior paraspinal edema ratio (SI-PER). The SI-PER was obtained by measuring the superior-inferior extent of increased signal/edema in the paraspinal soft tissues on the paraspinal inversion recovery images divided by the vertebral body height measured at midpoint. Cases positive for spondylodiscitis were those confirmed by biopsy, aspiration/drainage, surgery, or clinical improvement following antibiotic treatment. The diagnostic sensitivity and specificity of SI-PER were determined by Receiver operating characteristic (ROC) analysis. RESULTS: In 23/35 (66%) patients, the diagnosis of discitis/osteomyelitis was confirmed. The SI-PER showed a significant association with a positive MRI diagnosis (p = 0.001). Inter-observer correlation for SI-PER was 0.92. ROC analysis showed an area under the curve of 0.84. A SI-PER of 2.5 was 96% sensitive and 75% specific for the diagnosis of discitis/osteomyelitis, with a PPV of 88% and a NPV of 90%. CONCLUSION: In this study, the superior inferior paraspinal edema ratio (SI-PER), a newly defined MRI marker, was found to have high sensitivity for differentiating spondylodiscitis from endplate degenerative changes on lumbar spine MRI.

AutoProstate: Towards Automated Reporting of Prostate MRI for Prostate Cancer Assessment Using Deep Learning.

Multiparametric magnetic resonance imaging (mpMRI) of the prostate is used by radiologists to identify, score, and stage abnormalities that may correspond to clinically significant prostate cancer (CSPCa). Automatic assessment of prostate mpMRI using artificial intelligence algorithms may facilitate a reduction in missed cancers and unnecessary biopsies, an increase in inter-observer agreement between radiologists, and an improvement in reporting quality. In this work, we introduce AutoProstate, a deep learning-powered framework for automatic MRI-based prostate cancer assessment. AutoProstate comprises of three modules: Zone-Segmenter, CSPCa-Segmenter, and Report-Generator. Zone-Segmenter segments the prostatic zones on T2-weighted imaging, CSPCa-Segmenter detects and segments CSPCa lesions using biparametric MRI, and Report-Generator generates an automatic web-based report containing four sections: Patient Details, Prostate Size and PSA Density, Clinically Significant Lesion Candidates, and Findings Summary. In our experiment, AutoProstate was trained using the publicly available PROSTATEx dataset, and externally validated using the PICTURE dataset. Moreover, the performance of AutoProstate was compared to the performance of an experienced radiologist who prospectively read PICTURE dataset cases. In comparison to the radiologist, AutoProstate showed statistically significant improvements in prostate volume and prostate-specific antigen density estimation. Furthermore, AutoProstate matched the CSPCa lesion detection sensitivity of the radiologist, which is paramount, but produced more false positive detections.

Artificial Intelligence Compared to Radiologists for the Initial Diagnosis of Prostate Cancer on Magnetic Resonance Imaging: A Systematic Review and Recommendations for Future Studies.

Computer-aided diagnosis (CAD) of prostate cancer on multiparametric magnetic resonance imaging (mpMRI), using artificial intelligence (AI), may reduce missed cancers and unnecessary biopsies, increase inter-observer agreement between radiologists, and alleviate pressures caused by rising case incidence and a shortage of specialist radiologists to read prostate mpMRI. However, well-designed evaluation studies are required to prove efficacy above current clinical practice. A systematic search of the MEDLINE, EMBASE, and arXiv electronic databases was conducted for studies that compared CAD for prostate cancer detection or classification on MRI against radiologist interpretation and a histopathological reference standard, in treatment-naïve men with a clinical suspicion of prostate cancer. Twenty-seven studies were included in the final analysis. Due to substantial heterogeneities in the included studies, a narrative synthesis is presented. Several studies reported superior diagnostic accuracy for CAD over radiologist interpretation on small, internal patient datasets, though this was not observed in the few studies that performed evaluation using external patient data. Our review found insufficient evidence to suggest the clinical deployment of artificial intelligence algorithms at present. Further work is needed to develop and enforce methodological standards, promote access to large diverse datasets, and conduct prospective evaluations before clinical adoption can be considered.

Computer-aided diagnosis of prostate cancer using multiparametric MRI and clinical features: A patient-level classification framework.

Computer-aided diagnosis (CAD) of prostate cancer (PCa) using multiparametric magnetic resonance imaging (mpMRI) is actively being investigated as a means to provide clinical decision support to radiologists. Typically, these systems are trained using lesion annotations. However, lesion annotations are expensive to obtain and inadequate for characterizing certain tumor types e.g. diffuse tumors and MRI invisible tumors. In this work, we introduce a novel patient-level classification framework, denoted PCF, that is trained using patient-level labels only. In PCF, features are extracted from three-dimensional mpMRI and derived parameter maps using convolutional neural networks and subsequently, combined with clinical features by a multi-classifier support vector machine scheme. The output of PCF is a probability value that indicates whether a patient is harboring clinically significant PCa (Gleason score ≥3+4) or not. PCF achieved mean area under the receiver operating characteristic curves of 0.79 and 0.86 on the PICTURE and PROSTATEx datasets respectively, using five-fold cross-validation. Clinical evaluation over a temporally separated PICTURE dataset cohort demonstrated comparable sensitivity and specificity to an experienced radiologist. We envision PCF finding most utility as a second reader during routine diagnosis or as a triage tool to identify low-risk patients who do not require a clinical read.

Role of C-reactive protein in effective utilization of emergent MRI for spinal infections.

PURPOSE: Emergent spinal MRI is recommended for patients with back pain and red flags for infection. However, many of these studies are negative due to low prevalence of spinal infections. Our purpose was to assess if C-reactive protein (CRP) can be used to guide effective utilization of emergent MRI for spinal infections. METHODS: 316/960 (33%) MRIs performed for infection by the emergency department over 75-month period had CRP levels obtained at presentation, after excluding patients receiving antibiotic or had spinal surgery in < 1 month. An MRI was considered positive when there was imaging evidence of spinal infection confirmed on follow-up by surgery/biopsy/drainage or definitive therapy. A CRP of ≤ 10 mg/L was considered normal and > 100 mg/L as highly elevated. RESULTS: CRP was normal in 95/316 (30%) and abnormal in 221/316 (70%) patients. MRI was positive in 43/316 (13.6%) patients, all of whom had abnormal CRP. CRP (p < 0.001) and intravenous drug use (IVDU; p = 0.002) were independently associated with a positive MRI. Receiver operator characteristic (ROC) analysis showed AUC of 0.76 for CRP, slightly improving with IVDU. Sensitivity, specificity, and negative predictive values for CRP level cut-off: 10 mg/L, 100%, 35%, and 100%, and 100 mg/L, 58%, 70% and 91%, respectively. CONCLUSION: Abnormal CRP, although extremely sensitive, lacks specificity in predicting a positive MRI for spinal infection unless highly elevated. However, a normal CRP (absent recent antibiotic or surgery) makes spinal infection unlikely, and its routine use as a screening test can help reducing utilization of emergent MRI for this purpose.

Breast-Conserving Surgery Alone for Ductal Carcinoma In Situ: Factors Associated with Increased Risk of Local Recurrence.

PURPOSE: This retrospective study was aimed at identifying clinicopathologic characteristics associated with an increased risk for ipsilateral local recurrence (LR) in patients with ductal carcinoma in situ (DCIS) treated with wide local excision (WLE) alone without radiotherapy (RT). METHODS: All patients with DCIS treated with WLE alone at the Beth Israel Deaconess Medical Center, Boston, MA, USA, between the years 2000 and 2010 were identified. We collected data on demographics, parity, personal or family history of breast cancer, exogenous hormone use, tobacco use, comorbidities, genetic mutation carrier status, imaging interval, and tumor-specific characteristics. RESULTS: Overall, 222 patients were included in the study. Median follow-up time was 8 years. LR occurred in 9% of patients, with a recurrence rate of 11.3 per 1000 person-years. The risk of recurrence was lower for patients with nuclear grade (NG) I tumors than for patients with NG II or NG III tumors (3, 8.5, and 19%, respectively; p = 0.01). The median margin width was 1 mm in patients experiencing LR versus 1.8 mm in patients without LR (p = 0.3). Patients who had used exogenous hormones, or patients with a history of tobacco use, had higher rates of LR than those who did not, although the difference did not reach statistical significance. CONCLUSIONS: Our data indicate that higher NG, narrower margin width, use of exogenous hormones, and smoking history may be associated with an increased risk of LR. The evaluation of these factors may be helpful when considering whether or not to use adjuvant RT for patients with DCIS.

To stimulate or withdraw? A cost-utility analysis of recombinant human thyrotropin versus thyroxine withdrawal for radioiodine ablation in patients with low-risk differentiated thyroid cancer in the United States.

CONTEXT: Use of recombinant human TSH (rhTSH) prior to radioactive iodine remnant ablation for patients with differentiated thyroid cancer avoids the hypothyroid state and improves quality of life. European studies have shown that use of rhTSH vs. thyroid hormone withdrawal is a cost-effective method for preparing patients for ablation. OBJECTIVE: The objective of the study was to determine the cost-utility of rhTSH prior to ablation in the United States. DESIGN/SETTING/SUBJECTS: A Markov decision model was developed for a hypothetical group of adult patients with low-risk differentiated thyroid cancer who were prepared for ablation by either rhTSH or thyroid hormone withdrawal. Patients entered the model after initial thyroidectomy; follow-up was in accordance with current American Thyroid Association guidelines. Input data were obtained from the literature, Medicare reimbursement schedule, and U.S. Bureau of Labor Statistics. Sensitivity analyses were performed for all clinically relevant inputs. MAIN OUTCOME MEASURES: Cost-utility, measured in U.S. dollars per quality-adjusted life-year ($/QALY), was measured. RESULTS: Use of rhTSH yielded an incremental cost-utility of $52,554/QALY (95% confidence interval $52,058-53,050/QALY) (incremental societal cost of $1,365/patient; incremental benefit of 0.026 QALY/patient). The majority of cost and benefit occurs during the preablation, ablation, and postablation period; differences in cost are due to cost of rhTSH and differences in productivity loss (days off work). The model was most sensitive to changes in time off work, cost of rhTSH, and differences in utilities of health states. CONCLUSIONS: In the United States, the cost-effectiveness of rhTSH for ablation in patients with low-risk differentiated thyroid cancer is highly dependent on potential variations in cost of rhTSH, rates of remnant ablation, time off work, and quality of life.

Evaluating the surgery literature: can standardizing peer-review today predict manuscript impact tomorrow?

OBJECTIVE: Evidence-based surgery is predicated on the quality of published literature. We measured the quality of surgery manuscripts selected by peer review and identified predictors of excellence. METHODS: One hundred twenty clinical surgery manuscripts were randomly selected from 1998 in 5 eminent peer-reviewed surgery and medical journals. Manuscripts were blinded for author, institution, and journal of origin. Four surgeons and 4 methodologists evaluated the quality using novel instruments based on subject selection, study protocol, statistical analysis/inference, intervention description, outcome assessments, and results presentation. Predictors of quality and impact factor were identified using bivariate and multivariate regression. RESULTS: Oncology was the most common subject (26%), followed by general surgery/gastrointestinal (24%). The average number of study subjects was 417; the majority of manuscripts were American (53%), from a single institution (59%). Eighteen percent had a statistician author. Mean number of citations was 128. Surgery manuscripts from medical, compared with surgery journals, had better total quality scores (3.8 vs. 5.2, P < 0.001). They had more subjects and were more likely to have a statistician as coauthor (43% vs. 10%, P < 0.001), multi-institutional, international collaboration (30% vs. 8%, P < 0.001), and higher citation index (mean: 350 vs. 54, P < 0.001). They were more often foreign (70% vs. 40%, P < 0.001). Independent predictors of quality were having a statistician coauthor, study funding, European origin, and more study subjects. Quality assessment using our instruments predicted the number of citations after 10 years (P < 0.01), along with having a statistician coauthor, international multi-institutional collaboration, and more subjects. CONCLUSION: The quality of surgery manuscripts can be improved by including a statistician as coauthor, with efforts directed toward implementing multi-institutional/interdisciplinary trials. Peer-review across journals can be standardized through the use of instruments measuring methodologic and clinical quality.

Medullary thyroid cancer: early detection and novel treatments.

PURPOSE OF REVIEW: Medullary thyroid cancer (MTC) is derived from the parafollicular cells of the thyroid. Understanding the molecular biology behind specific mutations of the RET gene and their prognostic implications have led to the establishment of tailored treatment modalities for certain patients. We review the most recent studies on the molecular biology, calcitonin screening, diagnosis, imaging, and treatment of MTC. RECENT FINDINGS: Newly identified rearranged during transfection point mutations have helped with MTC prognosis and have resulted in the establishment of new treatment guidelines. Screening for MTC in the United States with basal serum calcitonin for patients with thyroid nodules would cost $11,793 per life-year saved (LYS), compared with colonoscopy and mammography screening. For metastatic or recurrent disease, neck ultrasound, chest computed tomography scan, liver MRI, bone scintigraphy, and axial skeleton MRI have been proven superior to 18F-FDG PET/computed tomography. For patients with nonoperable metastatic disease, novel chemotherapeutic agents, such as vandetanib, targeting rearranged during transfection, vascular endothelial growth factor receptor and epidermal growth factor receptor, are showing promise. Such agents are currently in phase II trials. SUMMARY: There have been several recent advances in the diagnosis, molecular biology, imaging, and treatment options of MTC. By potentially downstaging of disease, and treating metastatic disease more effectively, overall survival and outcomes of patients may improve.

The maturation of a specialty: Workforce projections for endocrine surgery.

BACKGROUND: There has been an increase in the incidence of endocrine diseases and the number of endocrine procedures in the United States. Higher surgeon volume is associated with improved patient outcomes. Fellowship programs will lead to more specialty-trained endocrine surgeons. We make projections for the supply of endocrine surgeons and demand for endocrine procedures over the next 15 years. METHODS: Supply projections are based on data from the Accreditation Council for Graduate Medical Education, a survey of American Association of Endocrine Surgery fellowship program graduates, and Healthcare Cost and Utilization Project National Inpatient Sample (HCUP-NIS). Demand is estimated using HCUP-NIS, U.S. Census Bureau projections, and a literature review. RESULTS: There were 64,275 endocrine procedures performed in 2000 and 80,505 in 2004. Using age-adjusted population projections and increasing incidence of endocrine diseases, 103,704 endocrine procedures are anticipated in 2020. High-volume endocrine surgeons are few in number, but perform 24% of endocrine procedures. Surgeon supply is projected to increase to 938 by 2020; this is based on fellowship graduation rates, retirement trends, and increasing annual endocrine case volume among high-volume surgeons. Alternative projections of supply and demand are generated to test the sensitivity of our analyses to different assumptions. CONCLUSION: Labor force planning in endocrine surgery is essential if the demand for more high-volume endocrine specialists is to be met.

Racial disparities in clinical and economic outcomes from thyroidectomy.

CONTEXT: Thyroid disease is common, and thyroidectomy is a mainstay of treatment for many benign and malignant thyroid conditions. Overall, thyroidectomy is associated with favorable outcomes, particularly if experienced surgeons perform it. OBJECTIVE: To examine racial differences in clinical and economic outcomes of patients undergoing thyroidectomy in the United States. DESIGN, SETTING, PATIENTS: The nationwide inpatient sample was used to identify thyroidectomy admissions from 1999 to 2004, using ICD-9 procedure codes. Race and other clinical and demographic characteristics of patients were collected along with surgeon volume and hospital characteristics to predict outcomes. MAIN OUTCOME MEASURES: Inpatient mortality, complication rates, length of stay (LOS), discharge status, and mean total costs by racial group. RESULTS: In 2003-2004, 16,878 patients underwent thyroid procedures; 71% were white, 14% black, 9% Hispanic, and 6% other. Mean LOS was longer for blacks (2.5 days) than for whites (1.8 days, P < 0.001); Hispanics had an intermediate LOS (2.2 days). Although rare, in-hospital mortality was higher for blacks (0.4%) compared with that for other races (0.1%, P < 0.001). Blacks trended toward higher overall complication rates (4.9%) compared with whites (3.8%) and Hispanics (3.6%, P = 0.056). Mean total costs were significantly lower for whites ($5447/patient) compared with those for blacks ($6587) and Hispanics ($6294). The majority of Hispanics (55%) and blacks (52%) had surgery by the lowest-volume surgeons (1-9 cases per year), compared with only 44% of whites. Highest-volume surgeons (>100 cases per year) performed 5% of thyroidectomies, but 90% of their patients were white (P < 0.001). Racial disparities in outcomes persist after adjustment for surgeon volume group. CONCLUSIONS: These findings suggest that, although thyroidectomy is considered safe, significant racial disparities exist in clinical and economic outcomes. In part, inequalities result from racial differences in access to experienced surgeons; more data are needed with regard to racial differences in thyroid biology and surveillance to explain the balance of observed disparities.