Synergizing Deep Learning-Enabled Preprocessing and Human-AI Integration for Efficient Automatic Ground Truth Generation.

The progress of incorporating deep learning in the field of medical image interpretation has been greatly hindered due to the tremendous cost and time associated with generating ground truth for supervised machine learning, alongside concerns about the inconsistent quality of images acquired. Active learning offers a potential solution to these problems of expanding dataset ground truth by algorithmically choosing the most informative samples for ground truth labeling. Still, this effort incurs the costs of human labeling, which needs minimization. Furthermore, automatic labeling approaches employing active learning often exhibit overfitting tendencies while selecting samples closely aligned with the training set distribution and excluding out-of-distribution samples, which could potentially improve the model's effectiveness. We propose that the majority of out-of-distribution instances can be attributed to inconsistent cross images. Since the FDA approved the first whole-slide image system for medical diagnosis in 2017, whole-slide images have provided enriched critical information to advance the field of automated histopathology. Here, we exemplify the benefits of a novel deep learning strategy that utilizes high-resolution whole-slide microscopic images. We quantitatively assess and visually highlight the inconsistencies within the whole-slide image dataset employed in this study. Accordingly, we introduce a deep learning-based preprocessing algorithm designed to normalize unknown samples to the training set distribution, effectively mitigating the overfitting issue. Consequently, our approach significantly increases the amount of automatic region-of-interest ground truth labeling on high-resolution whole-slide images using active deep learning. We accept 92% of the automatic labels generated for our unlabeled data cohort, expanding the labeled dataset by 845%. Additionally, we demonstrate expert time savings of 96% relative to manual expert ground-truth labeling.

Validation of the Foot and Ankle Outcome Score for Use in Infracalcaneal Heel Pain.

There is growing interest in adopting validated and reliable patient-reported outcome measures following surgery. While the Foot and Ankle Outcome Score (FAOS) has previously been validated for use in multiple foot/ankle conditions, it has not yet been validated in patients with infracalcaneal heel pain. In this study we aimed to validate the FAOS by looking at 4 psychometric properties of the survey: construct validity, content validity, reliability, and responsiveness, using patients in our practice with a clinical diagnosis of plantar fasciitis. A total of 150 patients (mean age 49.7 ± 12.1 years [36 men and 114 women]) were included in one or more of the 4 components of this study. All FAOS subscales demonstrated adequate construct validity when compared with the physical health component of the 12-Item Short Form Health Survey (SF-12), and 2 out of 5 subscales demonstrated moderate correlation with the mental health component of SF-12 (all Spearman rho >0.3, and p values <0.05). Most FAOS subscales demonstrated content validity and were found to contain relevant questions from the patient's perspective. All 5 subscales demonstrated good test-retest reliability with intraclass correlation coefficients ≥ 0.827. Finally, 4 out of the 5 subscales (all but other symptoms) were responsive to change at a mean follow up of 12.2 months after surgery (p < .05). We conclude that the FAOS is a responsive, reliable, and valid instrument for use in infracalcaneal heel pain. We believe that due to its ease of use and broad applicability, the FAOS could be more widely adopted in foot/ankle practices as patient-centered healthcare delivery and research becomes increasingly prioritized in the US and abroad.

Impaired Foot Plantar Flexor Muscle Performance in Individuals With Plantar Heel Pain and Association With Foot Orthosis Use.

Study Design Controlled laboratory study. Background Plantar heel pain is one of the most common foot and ankle conditions seen in clinical practice, and many individuals continue to have persisting or recurrent pain after treatment. Impaired foot plantar flexor muscle performance is a factor that may contribute to limited treatment success, but reliable methods to identify impairments in individuals with plantar heel pain are needed. In addition, foot orthoses are commonly used to treat this condition, but the implications of orthosis use on muscle performance have not been assessed. Objectives To assess ankle plantar flexor and toe flexor muscle performance in individuals with plantar heel pain using clinically feasible measures and to examine the relationship between muscle performance and duration of foot orthosis use. Methods The rocker-board plantar flexion test (RBPFT) and modified paper grip test for the great toe (mPGTGT) and lesser toes (mPGTLT) were used to assess foot plantar flexor muscle performance in 27 individuals with plantar heel pain and compared to 27 individuals without foot pain who were matched according to age, sex, and body mass. Pain ratings were obtained before and during testing, and self-reported duration of foot orthosis use was recorded. Results Compared to the control group, individuals with plantar heel pain demonstrated lower performance on the RBPFT (P = .001), the mPGTGT (P = .022), and the mPGTLT (P = .037). Longer duration of foot orthosis use was moderately correlated to lower performance on the RBPFT (r = -0.52, P = .02), the mPGTGT (r = -0.54, P = .01), and the mPGTLT (r = -0.43, P = .03). Conclusion Ankle plantar flexor and toe flexor muscle performance was impaired in individuals with plantar heel pain and associated with longer duration of self-reported foot orthosis use. J Orthop Sports Phys Ther 2016;46(8):681-688. Epub 3 Jul 2016. doi:10.2519/jospt.2016.6482.