Development of a commercially viable machine learning product to automatically detect rotator cuff muscle pathology

项目总结 肩袖撕裂对于大量患者来说是非常有问题的，因此仍然是一个非常具有挑战性的问题。 临床问题。在60岁的人中，大约有20%到50%的人有已知的肩袖撕裂和患病率 只随着年龄的增长而增加。虽然肩袖的外科重建寻求改善肩部功能和 稳定性，手术结果的成功程度差别很大。这些截然不同的结果是 因为，在手术前，在目前的评估方法下，很难预测哪些患者将受益于 手术对抗那些不愿做手术的人。该项目的重点是开发独特的技术，以取代现有的 方法对能够大规模商业部署的肩袖进行快速、准确的评估。 从临床的角度来看，有重要的科学证据表明，过度的脂肪渗透和萎缩 肩袖肌肉会导致不良的结果，因为脂肪组织的存在限制了 肌腱重建后恢复和再生肌肉。虽然目前的临床实践利用 核磁共振成像(MRI)使用定性评分系统评估肩袖中的脂肪渗透， 以往的研究已经证实，定性评分与定量评分的相关性相对较低 脂肪渗透和萎缩的测量。纳入量化测量将显著改善 临床治疗决策。然而，在现有方法下，这种评估将需要大量资金 人工输入，因此在临床上不可行。一种快速准确的肩袖肌肉分割方法 量化脂肪渗透对于改善结果和减少不必要的手术至关重要。 这项提议旨在利用跳羚之前在机器学习图像方面的技术创新 开发一种能够快速、准确地评估肩袖肌肉萎缩的分割算法 定量和脂肪渗透。将开发该算法，使其最终可以无缝集成 进入当前的临床工作流程，从而不需要任何额外的临床医生时间，并且事实上很可能 大大缩短了这段时间。在目标1中，我们将开发并验证一个基于深度学习的自动算法 用于量化肩袖肌肉体积和脂肪渗透。在目标2中，我们将开发一个软件 将算法整合到临床工作流程中以支持决策过程的原型。完成 该第一阶段项目的第一阶段将产生一个原型产品，该原型产品在第二阶段期间将在多个 整形外科中心，支持510(K)申请市场许可。这一项目将显著改善 提高肩关节病理评估的准确性，促进肩关节的诊断和治疗 病理，改善昂贵的整形外科手术的结果，甚至有可能消除 不必要的手术，所有这些都将改善患者护理并降低相关成本。

PROJECT SUMMARY Rotator cuff tears are highly problematic for large patient populations, and therefore remain a very challenging clinical problem. Roughly 20% to 50% of those 60 years of age have a known rotator cuff tear and the prevalence only increases with age. While surgical reconstruction of the rotator cuff seeks to improve shoulder function and stability, the degrees of successful surgical outcomes vary significantly. These widely differing outcomes are because, pre-operatively, it is difficult under current evaluative methods to predict which patients will benefit from surgery versus those who will not. The focus of this project is to develop unique technology that replaces current methods to produce a rapid, accurate assessment of rotator cuffs capable of large-scale commercial deployment. From a clinical perspective, there is significant scientific evidence that excessive fat infiltration and atrophy of the rotator cuff muscles lead to poor outcomes because the presence of fatty tissue limits the ability for the muscle to recover and regenerate following tendon reconstruction. While current clinical practice utilizes magnetic resonance imaging (MRI) to evaluate fat infiltration in the rotator cuff using qualitative scoring systems, previous studies have established that qualitative scoring has a relatively low correlation with quantitative measures of fat infiltration and atrophy. Incorporating quantitative measurements would dramatically improve clinical treatment decision-making. However, such evaluation under existing methods would require substantial manual input and thus is not clinically viable. A fast and accurate method for segmenting the rotator cuff muscles and quantifying fat infiltration is essential for improving outcomes and reducing unnecessary surgeries. This proposal aims to leverage Springbok’s previous technological innovations in machine learning image segmentation to develop an algorithm capable of fast, accurate assessment of rotator cuff muscle atrophy quantification and fat infiltration. The algorithm will be developed so that it can ultimately be seamlessly integrated into the current clinical workflow, thereby not requiring any additional clinician time, and in fact is likely to materially reduce that time. In Aim 1, we will develop and validate a deep-learning-based automatic algorithm for quantification of rotator cuff muscle volumes and fatty infiltration. In Aim 2, we will develop a software prototype to incorporate the algorithm into clinical workflow to support the decision-making process. Completion of this Phase 1 project will lead to a prototype product that is ready for beta-testing during Phase II at multiple Orthopaedic centers, enabling a 510(k) application for market clearance. This project will significantly improve the accuracy of shoulder pathology assessments, thus advancing the diagnosis and treatment of shoulder pathologies, improving the outcomes of costly Orthopaedic procedures, and potentially even eliminating unnecessary procedures, all of which will improve patient care and lower the associated costs.