Deep Learning Assistance Closes the Accuracy Gap in Fracture Detection Across Clinician Types.

BACKGROUND: Missed fractures are the most common diagnostic errors in musculoskeletal imaging and can result in treatment delays and preventable morbidity. Deep learning, a subfield of artificial intelligence, can be used to accurately detect fractures by training algorithms to emulate the judgments of expert clinicians. Deep learning systems that detect fractures are often limited to specific anatomic regions and require regulatory approval to be used in practice. Once these hurdles are overcome, deep learning systems have the potential to improve clinician diagnostic accuracy and patient care. QUESTIONS/PURPOSES: This study aimed to evaluate whether a Food and Drug Administration-cleared deep learning system that identifies fractures in adult musculoskeletal radiographs would improve diagnostic accuracy for fracture detection across different types of clinicians. Specifically, this study asked: (1) What are the trends in musculoskeletal radiograph interpretation by different clinician types in the publicly available Medicare claims data? (2) Does the deep learning system improve clinician accuracy in diagnosing fractures on radiographs and, if so, is there a greater benefit for clinicians with limited training in musculoskeletal imaging? METHODS: We used the publicly available Medicare Part B Physician/Supplier Procedure Summary data provided by the Centers for Medicare & Medicaid Services to determine the trends in musculoskeletal radiograph interpretation by clinician type. In addition, we conducted a multiple-reader, multiple-case study to assess whether clinician accuracy in diagnosing fractures on radiographs was superior when aided by the deep learning system compared with when unaided. Twenty-four clinicians (radiologists, orthopaedic surgeons, physician assistants, primary care physicians, and emergency medicine physicians) with a median (range) of 16 years (2 to 37) of experience postresidency each assessed 175 unique musculoskeletal radiographic cases under aided and unaided conditions (4200 total case-physician pairs per condition). These cases were comprised of radiographs from 12 different anatomic regions (ankle, clavicle, elbow, femur, forearm, hip, humerus, knee, pelvis, shoulder, tibia and fibula, and wrist) and were randomly selected from 12 hospitals and healthcare centers. The gold standard for fracture diagnosis was the majority opinion of three US board-certified orthopaedic surgeons or radiologists who independently interpreted the case. The clinicians' diagnostic accuracy was determined by the area under the curve (AUC) of the receiver operating characteristic (ROC) curve, sensitivity, and specificity. Secondary analyses evaluated the fracture miss rate (1-sensitivity) by clinicians with and without extensive training in musculoskeletal imaging. RESULTS: Medicare claims data revealed that physician assistants showed the greatest increase in interpretation of musculoskeletal radiographs within the analyzed time period (2012 to 2018), although clinicians with extensive training in imaging (radiologists and orthopaedic surgeons) still interpreted the majority of the musculoskeletal radiographs. Clinicians aided by the deep learning system had higher accuracy diagnosing fractures in radiographs compared with when unaided (unaided AUC: 0.90 [95% CI 0.89 to 0.92]; aided AUC: 0.94 [95% CI 0.93 to 0.95]; difference in least square mean per the Dorfman, Berbaum, Metz model AUC: 0.04 [95% CI 0.01 to 0.07]; p < 0.01). Clinician sensitivity increased when aided compared with when unaided (aided: 90% [95% CI 88% to 92%]; unaided: 82% [95% CI 79% to 84%]), and specificity increased when aided compared with when unaided (aided: 92% [95% CI 91% to 93%]; unaided: 89% [95% CI 88% to 90%]). Clinicians with limited training in musculoskeletal imaging missed a higher percentage of fractures when unaided compared with radiologists (miss rate for clinicians with limited imaging training: 20% [95% CI 17% to 24%]; miss rate for radiologists: 14% [95% CI 9% to 19%]). However, when assisted by the deep learning system, clinicians with limited training in musculoskeletal imaging reduced their fracture miss rate, resulting in a similar miss rate to radiologists (miss rate for clinicians with limited imaging training: 9% [95% CI 7% to 12%]; miss rate for radiologists: 10% [95% CI 6% to 15%]). CONCLUSION: Clinicians were more accurate at diagnosing fractures when aided by the deep learning system, particularly those clinicians with limited training in musculoskeletal image interpretation. Reducing the number of missed fractures may allow for improved patient care and increased patient mobility. LEVEL OF EVIDENCE: Level III, diagnostic study.

Reevaluation of missed lung cancer with artificial intelligence.

Lung cancer is often missed on chest radiographs, despite chest radiography typically being the first imaging modality in the diagnosis pathway. We present a 46 year-old male with chest pain referred for chest X-ray, and initial interpretation reported no abnormality within the patient's lungs. The patient was discharged but returned 4 months later with persistent and worsening symptoms. At this time, chest X-ray was again performed and revealed an enlarging left perihilar mass with post-obstructive atelectasis in the left lower lobe. Follow-up chest computerized tomography scan confirmed lung cancer with post-obstructive atelectasis, and subsequent bronchoscopy-assisted biopsy confirmed squamous cell carcinoma. Retrospective analysis of the initial chest radiograph, which had reported normal findings, was performed with Chest-CAD, a Food and Drug Administration (FDA) cleared computer-assisted detection (CAD) software device that analyzes chest radiograph studies using artificial intelligence. The device highlighted the perihilar region of the left lung as suspicious. Additional information provided by artificial intelligence software holds promise to prevent missed detection of lung cancer on chest radiographs.

Deep neural network improves fracture detection by clinicians.

Suspected fractures are among the most common reasons for patients to visit emergency departments (EDs), and X-ray imaging is the primary diagnostic tool used by clinicians to assess patients for fractures. Missing a fracture in a radiograph often has severe consequences for patients, resulting in delayed treatment and poor recovery of function. Nevertheless, radiographs in emergency settings are often read out of necessity by emergency medicine clinicians who lack subspecialized expertise in orthopedics, and misdiagnosed fractures account for upward of four of every five reported diagnostic errors in certain EDs. In this work, we developed a deep neural network to detect and localize fractures in radiographs. We trained it to accurately emulate the expertise of 18 senior subspecialized orthopedic surgeons by having them annotate 135,409 radiographs. We then ran a controlled experiment with emergency medicine clinicians to evaluate their ability to detect fractures in wrist radiographs with and without the assistance of the deep learning model. The average clinician's sensitivity was 80.8% (95% CI, 76.7-84.1%) unaided and 91.5% (95% CI, 89.3-92.9%) aided, and specificity was 87.5% (95 CI, 85.3-89.5%) unaided and 93.9% (95% CI, 92.9-94.9%) aided. The average clinician experienced a relative reduction in misinterpretation rate of 47.0% (95% CI, 37.4-53.9%). The significant improvements in diagnostic accuracy that we observed in this study show that deep learning methods are a mechanism by which senior medical specialists can deliver their expertise to generalists on the front lines of medicine, thereby providing substantial improvements to patient care.

Elevated rates of transaminitis during ipilimumab therapy for metastatic melanoma.

Melanoma is the deadliest form of skin cancer. Ipilimumab, a novel immunotherapy, is the first treatment shown to improve survival in patients with metastatic melanoma in large randomized controlled studies. The most concerning side effects reported in clinical studies of ipilimumab fall into the category of immune-related adverse events, which include enterocolitis, dermatitis, thyroiditis, hepatitis, hypophysitis, uveitis, and others. During the course of routine clinical care at Mount Sinai Medical Center, frequent hepatotoxicity was noted when ipilimumab was administered at a dose of 3 mg/kg according to Food and Drug Administration (FDA) guidelines. To better characterize these adverse events, we conducted a retrospective review of the first 11 patients with metastatic melanoma treated with ipilimumab at the Mount Sinai Medical Center after FDA approval. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) elevation, as defined by the National Cancer Institute's Common Terminology Criteria for Adverse Events, each occurred in six of 11 cases (≥grade 1), a notably higher frequency than could be expected on the basis of the FDA licensing study where elevations were reported in 0.8 and 1.5% of patients for AST and ALT, respectively. Grade 3 elevations in AST occurred in three of 11 patients as compared with 0% in the licensing trial. All cases of transaminitis resolved when ipilimumab was temporarily withheld without administration of immunosuppressive medication. During routine clinical care of late-stage melanoma patients with ipilimumab, physicians should monitor patients closely for hepatotoxicity and be aware that toxicity rates may differ across populations during ipilimumab therapy.

Actein inhibits the Na+-K+-ATPase and enhances the growth inhibitory effect of digitoxin on human breast cancer cells.

The Na+K+-ATPase is a known target of cardiac glycosides such as digitoxin and ouabain. We determined that the enzyme also is a target of the structurally-related triterpene glycoside actein, present in the herb black cohosh. Actein's inhibition of Na+-K+-ATPase activity was less potent than that of digitoxin, but actein potentiated digitoxin's inhibitory effect on Na+-K+-ATPase activity and MDA-MB-453 breast cancer cell growth. We observed different degrees of signal amplification for the two compounds. Actein's inhibitory effect on ATPase activity was amplified 2-fold for cell growth inhibition, whereas digitoxin's signal was amplified 20-fold. Actein induced a biphasic response in proteins downstream of ATPase: low dose and short duration of treatment upregulated NF-kappaB promoter activity, p-ERK, p-Akt and cyclin D1 protein levels, whereas higher doses and longer exposure inhibited these activities. Actein and digitoxin may be a useful synergistic combination for cancer chemoprevention and/or therapy.