Development, Deployment, and Implementation of a Machine Learning Surgical Case Length Prediction Model and Prospective Evaluation.

OBJECTIVE: To implement a machine learning model using only the restricted data available at case creation time to predict surgical case length for multiple services at different locations. BACKGROUND: The operating room is one of the most expensive resources in a health system, estimated to cost $22 to $133 per minute and generate about 40% of hospital revenue. Accurate prediction of surgical case length is necessary for efficient scheduling and cost-effective utilization of the operating room and other resources. METHODS: We introduced a similarity cascade to capture the complexity of cases and surgeon influence on the case length and incorporated that into a gradient-boosting machine learning model. The model loss function was customized to improve the balance between over- and under-prediction of the case length. A production pipeline was created to seamlessly deploy and implement the model across our institution. RESULTS: The prospective analysis showed that the model output was gradually adopted by the schedulers and outperformed the scheduler-predicted case length from August to December 2022. In 33,815 surgical cases across outpatient and inpatient platforms, the operational implementation predicted 11.2% fewer underpredicted cases and 5.9% more cases within 20% of the actual case length compared with the schedulers and only overpredicted 5.3% more. The model assisted schedulers to predict 3.4% more cases within 20% of the actual case length and 4.3% fewer underpredicted cases. CONCLUSIONS: We created a unique framework that is being leveraged every day to predict surgical case length more accurately at case posting time and could be potentially utilized to deploy future machine learning models.

Creation of an Automated Fluorescence Guided Tumor Ablation System.

OBJECTIVE: Create a device that improves the identification and extent of resection at the interface between healthy and tumor tissue; ultimately, using this device would improve surgical outcomes for patients and increase survival. METHODS: We have created a contactless tumor removal system that utilizes endogenous fluorescence feedback to inform a laser ablation system to execute autonomous removal of phantom tumor tissue. RESULTS: This completely non-contact surgical system is capable of resecting the tumor boundary of a tissue phantom with an average root mean square error (RMSE) of approximately 1.55 mm and an average max error of approximately 2.15 mm. There is no difference in the performance of the system when changing the size of the internal tumor from 7.5-12.5 mm in diameter. DISCUSSION: Future research steps include creating a more intelligent spectral search strategy to increase the density of points around the resection boundary, and to develop a more sophisticated classifier to predict pathologic diagnosis and tissue subtypes located regionally around the tumor boundaries. We envision this device being used to resect the boundaries of tumors identified by exogenously delivered tumor-labeling fluorophores, such as fluorescein or 5-ALA, in addition to approaches relying on autofluorescence of endogenous fluorophores.

Crew Recovery and Contingency Planning for a Manned Stratospheric Balloon Flight - the StratEx Program.

The StratEx program used a self-contained space suit and balloon system to loft pilot Alan Eustace to a record-breaking altitude and skydive from 135,897 feet (41,422 m). After releasing from the balloon and a stabilized freefall, the pilot safely landed using a parachute system based on a modified tandem parachute rig. A custom spacesuit provided life support using a similar system to NASA's (National Aeronautics and Space Administration; Washington, DC USA) Extravehicular Mobility Unit. It also provided tracking, communications, and connection to the parachute system. A recovery support team, including at least two medical personnel and two spacesuit technicians, was charged with reaching the pilot within five minutes of touchdown to extract him from the suit and provide treatment for any injuries. The team had to track the flight at all times, be prepared to respond in case of premature release, and to operate in any terrain. Crew recovery operations were planned and tailored to anticipate outcomes during this novel event in a systematic fashion, through scenario and risk analysis, in order to minimize the probability and impact of injury. This analysis, detailed here, helped the team configure recovery assets, refine navigation and tracking systems, develop procedures, and conduct training. An extensive period of testing and practice culminated in three manned flights leading to a successful mission and setting the record for exit altitude, distance of fall with stabilizing device, and vertical speed with a stabilizing device. During this mission, recovery teams reached the landing spot within one minute, extracted the pilot, and confirmed that he was not injured. This strategy is presented as an approach to prehospital planning and care for improved safety during crew recovery in novel, extreme events. Menon AS , Jourdan D , Nusbaum DM , Garbino A , Buckland DM , Norton S , Clark JB , Antonsen EL . Crew recovery and contingency planning for a manned stratospheric balloon flight - the StratEx program. Prehosp Disaster Med. 2016;31(5):524-531.

Long-term biomechanical outcomes after Achilles tendon ruptures.

PURPOSE: The ideal treatment for Achilles tendon ruptures is still unknown. Biomechanical were correlated to radiological and clinical parameters to study outcomes. METHODS: In this retrospective, assessor-blinded multi-centre cohort study, 52 patients with unilateral Achilles tendon rupture were assessed, each at least 3 years after injury. Patients underwent open surgery, percutaneous surgery or non-surgical treatment of Achilles tendon rupture. Both legs underwent plantar pressure distribution and isokinetic measures. Demographic parameters, maximum calf circumference (MCC) and clinical scores (American Orthopaedic Foot and Ankle Society, Achilles tendon rupture score, Hannover) were also evaluated. Complications were not assessed. RESULTS: Peak plantar flexion torque (PPFT) was significantly weaker on the treated side compared to the untreated leg [80.4 ± 29.7 Nm (mean ± SD) vs. 92.1 ± 27.4 Nm, p < 0.0001]. PPFT and push-off force (POFF) were not different between treatment groups nor was there a leg difference in POFF alone. There was only a weak correlation of clinical scores and PPFT or POFF, respectively. MCC correlated significantly with both PPFT (R (2) = 0.21, p = 0.01) and POFF (R (2) = 0.29, p < 0.0001). POFF appeared to be a predictor of PPFT (R (2) = 0.31, p < 0.0001). Open surgery outperformed non-surgical treatment in terms of centre-of-pressure line (p = 0.007), torque per muscle volume (p = 0.04) and relative POFF per body weight (p = 0.02) and relative in side comparison (p = 0.03). CONCLUSIONS: Clinical scores do not predict biomechanical outcomes. Clinically measured MCC is a good predictor of PPFT and POFF and can easily be used in clinical practice. Relative POFF in side comparison as well as per body weight favours surgical treatment.

All-inside meniscal repair devices compared with their matched inside-out vertical mattress suture repair: introducing 10,000 and 100,000 loading cycles.

BACKGROUND: All-inside arthroscopic meniscal repairs are favored by most clinicians because of their lower complication rate and decreased morbidity compared with inside-out techniques. Until now, only 1000 cycles have been used for biomechanical testing. HYPOTHESIS: All-inside meniscal repairs will show inferior biomechanical response to cyclic loading (up to 100,000 cycles) and load-to-failure testing compared with inside-out suture controls. STUDY DESIGN: Controlled laboratory study. METHODS: Bucket-handle tears in 72 porcine menisci were repaired using the Omnispan and Fast-Fix 360 (all-inside devices) and Orthocord 2-0 and Ultrabraid 2-0 sutures (matched controls). Initial displacement, displacement after cyclic loading (100, 500, 1000, 2000, 5000, 10,000, and 100,000 cycles) between 5 and 20 N, ultimate load to failure, and mode of failure were recorded, as well as stiffness. RESULTS: Initial displacement and displacement after cyclic loading were not different between the groups. The Omnispan repair demonstrated the highest load-to-failure force (mean ± SD, 151.3 ± 21.5 N) and was significantly stronger than all the other constructs (Orthocord 2-0, 105.5 ± 20.4 N; Ultrabraid 2-0, 93.4 ± 22.5 N; Fast-Fix 360, 76.6 ± 14.2 N) (P < .0001 for all). The Orthocord vertical inside-out mattress repair was significantly stronger than the Fast-Fix 360 repair (P = .003). The Omnispan (30.8 ± 3.5 N/mm) showed significantly higher stiffness compared with the Ultrabraid 2-0 (22.9 ± 6.9 N/mm, P < .0001) and Fast-Fix 360 (23.7 ± 3.9 N/mm, P = .001). The predominant mode of failure was suture failure. CONCLUSION: All-inside meniscal devices show comparable biomechanical properties compared with inside-out suture repair in cyclic loading, even after 100,000 cycles. CLINICAL RELEVANCE: Eight to 10 weeks of rehabilitation might not pose a problem for all repairs in this worst-case scenario.

Long-term outcomes of muscle volume and Achilles tendon length after Achilles tendon ruptures.

PURPOSE: The best treatment for Achilles tendon (AT) ruptures remains controversial. Long-term follow-up with radiological and clinical measurements is needed. METHODS: In this retrospective multicentre cohort study, patients (n = 52) were assessed at a mean of 91 months follow-up after unilateral AT rupture treated by open, percutaneous or conservative (non-surgical) treatment. Demographic parameters, time off work, maximum calf circumference and clinical scores (ATRS, Hannover, AOFAS) were evaluated. Muscle volume and cross-sectional area of the calf and AT length were measured on MR images and were compared between groups and to each patient's healthy contralateral leg. RESULTS: Reduced muscle volume was found across all groups with a higher muscle volume in the conservative (729.9 ± 130.3 cm(3)) compared to the percutaneous group (675.9 ± 207.4 cm(3), p = 0.04). AT length was longer in the affected leg (198.4 ± 24.1 vs. 180.6 ± 25.0 mm, p < 0.0001) without difference in subgroup analysis. Clinically measured ankle dorsiflexion showed poor correlation with AT length (R (2) = 0.07, p = 0.008). Muscle volume strongly correlated with the cross-sectional area (R (2) = 0.6, p < 0.0001) but showed a weak correlation with the Hannover score (R (2) = 0.08, p = 0.048). Maximum calf circumference correlated with muscle volume (R (2) = 0.42, p < 0.0001). CONCLUSIONS: No significant difference between the treatment groups was found in muscle volume, AT length, clinical measures or days off work. Cross-sectional area and maximum calf circumference are cost-effective measurements and a good approximation of muscle volume and can thus be used in a clinical setting while clinical dorsiflexion should not be used.