Medical Physics Leadership Academy Journal Club (Leadership Club) program: Two-year review in building a community of leaders.

The American Association of Physicists in Medicine began the Medical Physics Leadership Academy Journal Club in the fall of 2020. The initiative was launched to provide a forum for medical physicists to learn about leadership topics using published material, discuss and reflect on the material, and consider incorporating the discussed skills into their professional practice. This report presents the framework for the MPLA Journal Club program, describes the lessons learned over the last 2 years, summarizes the data collected from attendees, and highlights the roadmap for the program moving forward.

Development and multi-institutional validation of a convolutional neural network to detect vertebral body mis-alignments in 2D x-ray setup images.

BACKGROUND: Misalignment to the incorrect vertebral body remains a rare but serious patient safety risk in image-guided radiotherapy (IGRT). PURPOSE: Our group has proposed that an automated image-review algorithm be inserted into the IGRT process as an interlock to detect off-by-one vertebral body errors. This study presents the development and multi-institutional validation of a convolutional neural network (CNN)-based approach for such an algorithm using patient image data from a planar stereoscopic x-ray IGRT system. METHODS: X-rays and digitally reconstructed radiographs (DRRs) were collected from 429 spine radiotherapy patients (1592 treatment fractions) treated at six institutions using a stereoscopic x-ray image guidance system. Clinically-applied, physician approved, alignments were used for true-negative, "no-error" cases. "Off-by-one vertebral body" errors were simulated by translating DRRs along the spinal column using a semi-automated method. A leave-one-institution-out approach was used to estimate model accuracy on data from unseen institutions as follows: All of the images from five of the institutions were used to train a CNN model from scratch using a fixed network architecture and hyper-parameters. The size of this training set ranged from 5700 to 9372 images, depending on exactly which five institutions were contributing data. The training set was randomized and split using a 75/25 split into the final training/ validation sets. X-ray/ DRR image pairs and the associated binary labels of "no-error" or "shift" were used as the model input. Model accuracy was evaluated using images from the sixth institution, which were left out of the training phase entirely. This test set ranged from 180 to 3852 images, again depending on which institution had been left out of the training phase. The trained model was used to classify the images from the test set as either "no-error" or "shifted", and the model predictions were compared to the ground truth labels to assess the model accuracy. This process was repeated until each institution's images had been used as the testing dataset. RESULTS: When the six models were used to classify unseen image pairs from the institution left out during training, the resulting receiver operating characteristic area under the curve values ranged from 0.976 to 0.998. With the specificity fixed at 99%, the corresponding sensitivities ranged from 61.9% to 99.2% (mean: 77.6%). With the specificity fixed at 95%, sensitivities ranged from 85.5% to 99.8% (mean: 92.9%). CONCLUSION: This study demonstrated the CNN-based vertebral body misalignment model is robust when applied to previously unseen test data from an outside institution, indicating that this proposed additional safeguard against misalignment is feasible.

AAPM task group report 275.S: Survey strategy and results on plan review and chart check practices in US and Canada.

BACKGROUND: AAPM Task Group (TG) 275 was charged with developing practical, evidence-based recommendations for physics plan and chart review clinical processes for radiation therapy. As part of this charge, and to characterize practices and clinical processes, a survey of the medical physics community was developed and conducted. Detailed analyses and trends based on the survey that exceeded TG report length constraints are presented herein. AIMS: The design, development, and detailed results of the TG- 275 survey as well as statistical analysis and trends are described in detail. This is complementary material to the TG 275 report. METHODS AND MATERIALS: The survey consisted of 100 multiple-choice questions divided into four main sections: 1) Demographics, 2) Initial Plan Check, 3) On-Treatment, and 4) End-of-Treatment Chart Check. The survey was released to all AAPM members who self-reported working in the radiation oncology field, and it was kept open for 7 weeks. Results were summarized using descriptive statistics. To study practice differences, tests of association were performed using data grouped by four demographic questions: 1) Institution Type, 2) Average number of patients treated daily, 3) Radiation Oncology Electronic Medical Record, and 4) Perceived Culture of Safety. RESULTS: The survey captured 1370 non-duplicate entries from the United States and Canada. Differences across practices were grouped and presented based on Process-Based and Check-Specific questions. A risk-based summary was created to show differences amongst the four demographic questions for checks associated with the highest risk failure modes identified by TG-275. CONCLUSION: The TG-275 survey captured a baseline of practices on initial plan, on-treatment, and end-of-treatment checks across a wide variety of clinics and institutions. The results of test of association showed practice heterogeneities as a function of demographic characteristics. Survey data were successfully used to inform TG-275 recommendations.

Phosphatidylinositol 3-kinase and Akt nonautonomously promote perineurial glial growth in Drosophila peripheral nerves.

Drosophila peripheral nerves, structured similarly to their mammalian counterparts, comprise a layer of motor and sensory axons wrapped by an inner peripheral glia (analogous to the mammalian Schwann cell) and an outer perineurial glia (analogous to the mammalian perineurium). Growth and proliferation within mammalian peripheral nerves are increased by Ras pathway activation: loss-of-function mutations in Nf1, which encodes the Ras inhibitor neurofibromin, cause the human genetic disorder neurofibromatosis, which is characterized by formation of neurofibromas (tumors of peripheral nerves). However, the signaling pathways that control nerve growth downstream of Ras remain incompletely characterized. Here we show that expression specifically within the Drosophila peripheral glia of the constitutively active Ras(V12) increases perineurial glial thickness. Using chromosomal loss-of-function mutations and transgenes encoding dominant-negative and constitutively active proteins, we show that this nonautonomous effect of Ras(V12) is mediated by the Ras effector phosphatidylinositol 3-kinase (PI3K) and its downstream kinase Akt. We also show that the nonautonomous, growth-promoting effects of activated PI3K are suppressed by coexpression within the peripheral glia of FOXO+ (forkhead box O) a transcription factor inhibited by Akt-dependent phosphorylation. We suggest that Ras-PI3K-Akt activity in the peripheral glia promotes growth of the perineurial glia by inhibiting FOXO. In mammalian peripheral nerves, the Schwann cell releases several growth factors that affect the proliferative properties of neighbors. Some of these factors are oversecreted in Nf1 mutants. Our results raise the possibility that neurofibroma formation in individuals with neurofibromatosis might result in part from a Ras-PI3K-Akt-dependent inhibition of FOXO within Schwann cells.