Histopathology-Based Diagnosis of Oral Squamous Cell Carcinoma Using Deep Learning

Oral squamous cell carcinoma (OSCC) is prevalent around the world and is associated with poor prognosis. OSCC is typically diagnosed from tissue biopsy sections by pathologists who rely on their empirical experience. Deep learning models may improve the accuracy and speed of image classification, thus reducing human error and workload. Here we developed a custom-made deep learning model to assist pathologists in detecting OSCC from histopathology images. We collected and analyzed a total of 2,025 images, among which 1,925 images were included in the training set and 100 images were included in the testing set. Our model was able to automatically evaluate these images and arrive at a diagnosis with a sensitivity of 0.98, specificity of 0.92, positive predictive value of 0.924, negative predictive value of 0.978, and F1 score of 0.951. Using a subset of 100 images, we examined whether our model could improve the diagnostic performance of junior and senior pathologists. We found that junior pathologists were able to delineate OSCC in these images 6.26 min faster when assisted by the model than when working alone. When the clinicians were assisted by the model, their average F1 score improved from 0.9221 to 0.9566 in the case of junior pathologists and from 0.9361 to 0.9463 in the case of senior pathologists. Our findings indicate that deep learning can improve the accuracy and speed of OSCC diagnosis from histopathology images.

Comparison of survival after mitral valve replacement with biologic and mechanical valves in 1139 patients

OBJECTIVE

We sought to compare 10-year survival in patients after mitral valve replacement with biologic or mechanical valve prostheses.

METHODS

Retrospective survival analysis was performed on data from 1139 consecutive patients older than 18 years of age undergoing mitral valve replacement with Carpentier-Edwards (n = 495; Baxter Healthcare Corp, Irvine, Calif) or St Jude Medical (n = 644; St Jude Medical, Inc, St Paul, Minn) prostheses.

RESULTS

The 10-year survival was not statistically different between the patients receiving Carpentier-Edwards valves and those receiving St Jude Medical valves (P =.16). Adjusted survival estimates at 2, 5, and 10 years were 82% +/- 2% (95% confidence intervals, 79%-85%), 69% +/- 2% (95% confidence intervals, 64%-73%), and 42% +/- 3% (95% confidence intervals, 37%-48%), respectively, for the Carpentier-Edwards group and 83% +/- 2% (95% confidence intervals, 80%-86%), 72% +/- 2% (95% confidence intervals, 69%-76%), and 51% +/- 3% (95% confidence intervals, 45%-58%), respectively, for the St Jude Medical group. Predictors of worse survival after mitral valve replacement are older age, lower ejection fraction, presence of class IV congestive heart failure, coronary artery disease, renal disease, smoking history, hypertension, concurrent other valve surgery, and redo heart surgery.

CONCLUSION

Choice of biologic or mechanical prosthesis does not significantly affect long-term patient survival after mitral valve replacement.

Long-term outcome after biologic versus mechanical aortic valve replacement in 841 patients

OBJECTIVE

The purpose of this study was to optimize selection criteria of biologic versus mechanical valve prostheses for aortic valve replacement.

METHODS

Retrospective analysis was performed for 841 patients undergoing isolated, first-time aortic valve replacement with Carpentier-Edwards (n = 429) or St Jude Medical (n = 412) prostheses.

RESULTS

Patients with Carpentier-Edwards and St Jude Medical valves had similar characteristics. Ten-year survival was similar in each group (Carpentier-Edwards 54% 3% versus St Jude Medical 50% 6%; P =.4). Independent predictors of worse survival were older age, renal or lung disease, ejection fraction less than 40%, diabetes, and coronary disease. Carpentier-Edwards versus St Jude Medical prostheses did not affect survival (P =.4). Independent predictors of aortic valve reoperation were younger age and Carpentier-Edwards prosthesis. The linearized rates of thromboembolism were similar, but the linearized rate of hemorrhage was lower with Carpentier-Edwards prostheses (P <.01). Perivalvular leak within 6 months of operation was more likely with St Jude Medical than with Carpentier-Edwards prostheses (P =.02). Estimated 10-year survival free from valve-related morbidity was better for the St Jude Medical valve in patients aged less than 65 years and was better for the Carpentier-Edwards valve in patients aged more than 65 years. Patients with renal disease, lung disease (in patients more than age 60 years), ejection fraction less than 40%, or coronary disease had a life expectancy of less than 10 years.

CONCLUSIONS

For first-time, isolated aortic valve replacement, mechanical prostheses should be considered in patients under age 65 years with a life expectancy of at least 10 years. Bioprostheses should be considered in patients over age 65 years or with lung disease (in patients over age 60 years), renal disease, coronary disease, ejection fraction less than 40%, or a life expectancy less than 10 years.

Determinants of 15-year outcome with 1,119 standard Carpentier-Edwards porcine valves

BACKGROUND

The determinants of long-term outcome 15 years or more after porcine valve replacement are poorly documented.

METHODS

A retrospective review was performed of patients undergoing valve replacement with standard Carpentier-Edwards aortic (n = 531), mitral (n = 492), and tricuspid (n = 96) valves.

RESULTS

Patient survival was 26%+/-3%, 23%+/-2%, and 31%+/-8% 15 years after aortic, mitral, and tricuspid valve replacements, respectively. Independent determinants of impaired long-term survival for aortic or mitral valve replacement were multiple valve replacement, older age, renal disease, lung disease, or coronary disease. Actual (versus actuarial) freedom from reoperation at 15 years was 86%+/-2%, 76%+/-2%, and 95%+/-2% after aortic, mitral, and tricuspid valve replacement, respectively. Risk factors for reoperation were young age for aortic or mitral valve replacement, previous operation for aortic valve replacement, and large valve size for mitral valve replacement. Freedom from thromboembolism was 77%+/-4%, 62%+/-9%, and 80%+/-5%; from hemorrhage, 95%+/-5%, 87%+/-4%, and 82%+/-6%; and from endocarditis, 94%+/-1%, 96%+/-1%, and 89%+/-5% 15 years after aortic, mitral, and tricuspid valve replacement, respectively. Risk factors for thromboembolism or hemorrhage were multiple valve replacement and age.

CONCLUSIONS

The standard Carpentier-Edwards bioprosthesis continues to provide relatively low complication rates at 15 years, especially in the aortic and tricuspid positions, and especially in patients older than 60 years or with significant comorbdity.

Comparison of open mitral commissurotomy with mitral valve replacement with or without chordal preservation in patients with mitral stenosis

BACKGROUND

Percutaneous balloon mitral valvuloplasty and durable mitral prostheses have made the role of open mitral commissurotomy (OMC) uncertain.

METHODS AND RESULTS

Results from the use of St Jude mitral valve replacement (SJMVR) were compared with those of the use of OMC in 312 consecutive patients with mitral stenosis between 1983 and the present. OMC and SJMVR patients were well matched for age, sex, and comorbidity except that SJMVR patients had more severe stenosis and were more likely to undergo concurrent aortic valve replacement. Compared with OMC, SJMVR without chordal preservation involved a longer pump time (158 +/- 81 versus 87 +/- 41 min, P < 0.05), more frequent in-hospital complications or death (57 of 219 [26%] versus 4 of 52 [8%], P < 0.01), and longer hospital stay (13 +/- 11 versus 10 +/- 6 days, P = 0.001). Preservation of chordae to at least 1 mitral valve leaflet decreased early morbidity and mortality rates of SJMVR to values comparable to those of OMC (3 of 41 [7%]). Survival was greater at 10 years for OMC versus SJMVR (86 +/- 5% versus 67 +/- 4%, P = 0.03). Ten-year freedom from cardiac events was not different between groups (49 +/- 9% for OMC versus 55 +/- 4% for SJMVR, P = 0.7). Freedom from subsequent mitral procedures at 10 years was better for SJMVR (96 +/- 2% versus 58 +/- 8%, P < 0.001).

CONCLUSIONS

In the modern era, SJMVR offers significantly greater durability than does OMC. Chordal preservation at the time of SJMVR may reduce perioperative complications to levels comparable to those of OMC.