Posttraumatic Stress Disorder (PTSD) is a common psychiatric disorder associated with comorbid psychiatric and physical disorders. The diagnosis, prognosis, and assessment of response to treatment of patients with PTSD, are based on clinical evaluation, and currently, there are no objective, clinically useful biomarkers to guide precise, personalized clinical decision making.
A proof of concept (POC) study aimed at demonstrating the validity and feasibility of using machine learning(ML) methods to differentiate between the PTSD and control groups based on each participant’s personal digital phenotype.
A prospective non-clinical trial conducted in adult men and women from the PTSD clinic, Chaim Sheba Medical Center. Cases (n=14) include participants with confirmed PTSD. The concurrent control group (n=15) includes participants without PTSD. A smart watch was used to measure the biomarkers and to communicate them to a smartphone. A decision tree with a 29-Fold cross validation method was performed to differentiate persons with PTSD from non-PTSD. Diagnostic performance metrics were calculated and compared to a random analysis.
We identified a statistically significant difference in eight of the eleven variables analyzed in the “Absolute change” and achieved greater diagnostic performance metrics than random when a stressor stimulus was present.
This study shows the feasibility of using digital health technology (DHT) in people with PTSD by finding objective measures that can be clinically useful in psychiatry.
Jessica Eve Bendheim MD MPH, Sean Zadik MD, Torr Polakow PhD, Eyal Zimlichman MD MSc, Raz Gross MD MPH and Nadav Goldental MD, MBA
6 – 14
Osteosarcoma is the most common malignant bone tumor in children. While limb-salvage surgery (LSS) is preferred over amputation, reconstruction remains especially challenging in pediatric patients due to high complication rates associated with traditional implants, including stress shielding and implant failure.
A 3-year-old female patient diagnosed with osteoblastic osteosarcoma of the right proximal femur underwent tumor resection and reconstruction with a custom prosthesis and bone allograft, which later failed. Revision surgery utilized a 3D-printed Ti-6Al-4V lattice implant with a porous structure for osteointegration and an inner tunnel for intramedullary fixation. An auto-graft was added to enhance bone regeneration.
This case demonstrates the first use of a 3D-printed femoral lattice implant in a pediatric osteosarcoma patient, offering a promising alternative to conventional implants. While early outcomes are positive, long-term follow-up is needed to assess its adaptation to skeletal growth.
Sana Zahalka, Yoav Levy, Liat Tsoren, Amit Benady MD PhD, Eran Golden, Yair Gortzak MD MSc, Solomon Dadia MD
15 – 19
