Evaluation of a Deep Image to Image Network (DI2IN) Auto-Segmentation Algorithm across a Network of Cancer Centers

PURPOSE/OBJECTIVE(S)

Due to manual OAR contouring challenges, various automatic contouring solutions have been introduced. Historically, common clinical auto-segmentation algorithms used were atlas based, which required maintaining a library of self-made contours. Searching the collection was computationally intensive and could take several minutes to complete. Deep learning approaches have shown significant benefits compared to atlas-based methods in improving segmentation accuracy and efficiency in auto-segmentation algorithms. This work represents the first multi-institutional study to describe and evaluate an AI algorithm for auto-segmentation of organs at risk (OARs) based on a deep image-to-image network (DI2IN).

MATERIALS/METHODS

The AI-Rad Companion Organs RT (AIRC) algorithm (Siemens Healthineers, Erlangen, Germany) uses a two-step approach for segmentation. In the first step, the target organ region in the optimal input image is extracted using a trained Deep Reinforcement Learning network (DRL), which is then used as input to create the contours in the second step based on DI2IN. The study was initially designed as a prospective single-center evaluation. The automated contours generated by AIRC were evaluated by three experienced board-certified radiation oncologists using a 4-point scale where 4 is clinically usable and 1 requires re-contouring. After seeing favorable results in a single-center pilot study, we decided to expand the study to 6 additional institutions, encompassing 8 additional evaluators for a total of 11 physician evaluators across 7 institutions.

RESULTS

One hundred fifty-six patients and 1366 contours were prospectively evaluated. The 5 most commonly contoured organs were the lung (136 contours, average rating = 4.0), spinal cord (106 contours, average rating = 3.1), eye globe (80 contours, average rating = 3.9), lens (77 contours, average rating = 3.9), and optic nerve (75 contours, average rating = 4.0). The average rating per evaluator per contour was 3.6. On average 124 contours were evaluated by each evaluator. 65% of the contours were rated as 4 and 31% were rated as 3. Only 4% of contours were rated as 1 or 2. 33 organs were evaluated in the study, with 19 structures having a 3.5 or above average rating (ribs, abdominopelvic cavity, skeleton, larynx, lung, aorta, brachial plexus, lens, eye globe, glottis, heart, parotid glands, bladder, kidneys, supraglottic larynx, submandibular glands, esophagus, optic nerve, oral cavity) and the remaining organs having a rating of 3.0 or greater (female breast, proximal femur, seminal vesicles, rectum, sternum, brainstem, prostate, brain, lips, mandible, liver, optic chiasm, spinal cord, spleen). No organ had an average rating below 3.

CONCLUSION

AIRC performed well with greater than 95% of contours accepted by treating physicians with no or minor edits. It supported a fully automated workflow with the potential for time savings and increased standardization with the use of AI-powered algorithms for high quality OAR contouring.

Pelvic Nodal Auto-Segmentation Using a Deep Image to Image Network (DI2IN) Auto-Segmentation Algorithm: Comparing Male vs. Female Pelvis

PURPOSE/OBJECTIVE(S)

Deep Learning approaches have shown significant benefits compared to atlas-based methods in improving segmentation accuracy and efficiency in auto-segmentation algorithms. The AI-Rad Companion Organs RT pelvic nodal auto-segmentation feature was trained and developed for use in the male pelvis. There is no real-world data on its usability in male pelvis and whether it can be used for female pelvic nodal anatomy. This work represents the first multi-institutional study to describe and evaluate an AI algorithm for auto-segmentation of the pelvic nodal region in female patients based on a deep image-to-image network (DI2IN).

MATERIALS/METHODS

The AIRC algorithm uses a two-step approach for segmentation. In the first step, the target organ region in the optimal input image is extracted using a trained Deep Reinforcement Learning network (DRL), which is then used as input to create the contours in the second step based on DI2IN. We retrospectively evaluated AIRC pelvic nodal auto-segmentation in both male and female patients treated at our network of institutions. The automated pelvic nodal contours generated by AIRC were evaluated by one board-certified radiation oncologist, specializing in prostate and gynecologic malignancies. A 4-point scale was used, where 4 is clinically usable and 1 requires re-contouring. Pelvic nodal regions included the right and left side of the common iliac, external iliac, internal iliac, obturator and midline presacral nodes. A chi-squared test was then used to compare the scores of male and female pelvic nodal cases.

RESULTS

Fifty-two female and 51 male patients were included in the study, representing a total of 468 and 447 pelvic nodal regions, respectively. 96% (450 pelvic nodal contours) and 99% (443 pelvic nodal contours) required no or minor edits for female and male patients, respectively (p = 0.004). The right internal iliac was the only nodal group with a statistically significant difference between female (92% requiring no or minor edits) and male (100% requiring no or minimal edits) patients, p = 0.04. The percentage of patients requiring no, or minor edits was 87% (45 patients) and 92% (47 patients) for female and male patients, respectively (p = 0.36).

CONCLUSION

AIRC pelvic nodal auto-segmentation performed very well in both male and female pelvic nodal regions, with the male pelvic nodal regions performing better especially in the right internal iliac nodal group. It is usable in female pelvic nodal regions, with 96% of contours requiring no or minor edits. As auto-segmentation becomes more widespread, it may be important to have equal representation from all genders in training and validation of auto-segmentation algorithms.

Pilot Study of Peer Review in Low Middle-Income Country (LMIC) through Cloud-Based Platform

PURPOSE/OBJECTIVE(S)

Peer review is an essential step in clinical quality assurance and can impact patient safety and treatment outcomes. Most published data on peer review is from developed nations, with little data on peer review from low middle income countries (LMIC). Major challenges to peer review can include a lack of time, expertise and commitment from team members. With increasing access to advanced technology in LMIC, peer review is becoming more important to maintain quality and standard of care. We evaluated cloud-based e- peer review (Varian) in our network of hospitals in India with an aim to see feasibility and impact on care.

MATERIALS/METHODS

Four of 15 centers across India were selected for this pilot study. All team members were trained on the platform prior to implementation. New cases for the week treated with definitive intent were selected by the dosimetrist. The link to the cases were sent through email to reviewing physicians. Various aspects which were reviewed for each case were.1) Work up & staging (Documents were scanned and loaded).2) Treatment intent & prescription.3) Target contours.4) Normal Organ at risk contours.5) Dose- Volume -Histogram (DVH) with clinical goals attached. Cases were marked as "Not Appropriate", "Appropriate", "Appropriate with minor finding", "Represent with major revisions" as per volume and plan review.

RESULTS

Over a period of 2 months, a total of 80 cases underwent e-Peer Review at our network of hospitals prior to the start of treatment. Median turnover time (like from link sent to time to completion of review) was 48 (6-360) hours. Mean time taken by physician for review was 9 minutes (range 3 to15). 31.2% of cases were accepted without any changes, 51.9 % had a minor change and 16.9 % cases had major changes. Most frequent reason of major changes was contouring corrections 16.9%. 31% of major changes underwent recontouring and replanning before initiation of treatment.

CONCLUSION

Peer review was feasible in our network through this e-peer review system, with average turnover time and mean time taken for review of 48 hours & 9 min respectively. Peer review led to significant changes which could impact patient care delivery and outcome. The ability to review cases asynchronously via this cloud-based e-peer review system, helped to ease the burden of scheduling between treating and reviewing physician. We plan to implement this across the remaining centers in our network.

P-055 LAPAROSCOPIC EXPLANTATION OF GROIN MESH WITH STAPLED EXCISION OF COLO CUTANEOUS FISTULA

Aim

Groin mesh explantation is a challenging situation and a multidisciplinary approach at index explantation surgery can avoid a long term catastrophic situation for the patient .

Material & Method

The history of the patient started a decade back with bilateral laparoscopic TAPP, following which he had swelling in both groin area with high grade fever . He was advised open bilateral groin exploration and explant of the mesh. Post procedure he was kept on antibiotics for few weeks and he was asymptomatic, he developed swelling again and started on anti mycobacterial treatment

He presented four years with foul smelling discharge from the suprapubic area . CT Scan with sinogram revealed a communication with sigmoid colon and tacker from index surgery

We planned laparoscopic approach, stapled excision of sigmoid-cutaneous fistula was performed . Removal of residual mesh and tacker from index under fluoroscopy guidance with skin tract excision .Intra op flexible sigmoidoscopy revealed no lumen compromise .

Histopathology showed granulomatous inflammation for which he was started on atypical mycobacterial treatment for 9 months .He remains uneventful after 2 years of surgery .

Conclusion

Mesh infection should be avoided at index surgery with utmost care to aseptic precautions.

Mesh explantation should be a part of multidisciplinary approach, radiology and colonoscopy were key investigations in this case . One should plan for a complete removal of mesh and tacker/s at index explantation surgery for safe long term outcomes .

Evaluation of prognostic risk models for postoperative pulmonary complications in adult patients undergoing major abdominal surgery: a systematic review and international external validation cohort study

BACKGROUND

Stratifying risk of postoperative pulmonary complications after major abdominal surgery allows clinicians to modify risk through targeted interventions and enhanced monitoring. In this study, we aimed to identify and validate prognostic models against a new consensus definition of postoperative pulmonary complications.

METHODS

We did a systematic review and international external validation cohort study. The systematic review was done in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. We searched MEDLINE and Embase on March 1, 2020, for articles published in English that reported on risk prediction models for postoperative pulmonary complications following abdominal surgery. External validation of existing models was done within a prospective international cohort study of adult patients (≥18 years) undergoing major abdominal surgery. Data were collected between Jan 1, 2019, and April 30, 2019, in the UK, Ireland, and Australia. Discriminative ability and prognostic accuracy summary statistics were compared between models for the 30-day postoperative pulmonary complication rate as defined by the Standardised Endpoints in Perioperative Medicine Core Outcome Measures in Perioperative and Anaesthetic Care (StEP-COMPAC). Model performance was compared using the area under the receiver operating characteristic curve (AUROCC).

FINDINGS

In total, we identified 2903 records from our literature search; of which, 2514 (86·6%) unique records were screened, 121 (4·8%) of 2514 full texts were assessed for eligibility, and 29 unique prognostic models were identified. Nine (31·0%) of 29 models had score development reported only, 19 (65·5%) had undergone internal validation, and only four (13·8%) had been externally validated. Data to validate six eligible models were collected in the international external validation cohort study. Data from 11 591 patients were available, with an overall postoperative pulmonary complication rate of 7·8% (n=903). None of the six models showed good discrimination (defined as AUROCC ≥0·70) for identifying postoperative pulmonary complications, with the Assess Respiratory Risk in Surgical Patients in Catalonia score showing the best discrimination (AUROCC 0·700 [95% CI 0·683-0·717]).

INTERPRETATION

In the pre-COVID-19 pandemic data, variability in the risk of pulmonary complications (StEP-COMPAC definition) following major abdominal surgery was poorly described by existing prognostication tools. To improve surgical safety during the COVID-19 pandemic recovery and beyond, novel risk stratification tools are required.

FUNDING

British Journal of Surgery Society.

Patient perceptions of psoriatic disease in Japan: Results from the Japanese subgroup of the Understanding Psoriatic Disease Leveraging Insights for Treatment (UPLIFT) survey

The population‐based Understanding Psoriatic Disease Leveraging Insights for Treatment (UPLIFT) survey was designed to better understand patient and dermatologist perceptions of the disease burden of psoriasis (PsO) and their treatment expectations. UPLIFT was a cross‐sectional, quantitative, online survey conducted in Europe, North America, and Japan between 2 March and 3 June 2020. In Japan, 391 patients reporting a diagnosis of PsO and/or psoriatic arthritis (PsA) were surveyed (75% had PsO alone, 23% had PsO and PsA, and 2% had PsA alone). Self‐reported body surface area (BSA) data were available for 309 Japanese patients, with the majority (80%) reporting PsO‐involved BSA ≤3 palms. Current symptoms of PsO were rated as moderate or severe by 43% of Japanese patients with BSA ≤3 palms, and severe by 44% of patients with BSA 4–10 palms. PsO frequently occurred in ≥1 special areas, most commonly the scalp in 76% of Japanese patients with BSA ≤3 palms, and ≥90% of those with BSA ≥4 palms. Furthermore, musculoskeletal symptoms in 42% of patients with PsO alone were suggestive of PsA. Whereas Japanese patients with BSA ≤3 palms mainly reported receiving topical therapy alone (34%) or no treatment (32%), 64% patients with BSA 4–10 palms reported receiving systemic therapy. Overall, 21% of Japanese patients with self‐perceived mild symptoms of PsO and 48% of patients with special area involvement experienced at least a moderate impact of disease on quality of life (Dermatology Life Quality Index score >5). Moreover, patients and dermatologists differed in their perceptions of determinants of PsO severity and treatment, and office visit discussions. In general, these findings from the Japanese subgroup of the UPLIFT survey demonstrated that a high proportion of patients perceived their symptoms to be moderate or severe irrespective of the level of skin involvement, suggesting a persistent unmet treatment need.

Chronic unilateral cervical vagotomy reduces renal inflammation, blood pressure, and renal injury in a mouse model of lupus

Systemic lupus erythematosus (SLE) is characterized by hypertension that results from chronic renal inflammation and dysautonomia in the form of dampened vagal tone. In health, the vagus nerve regulates inflammatory processes through mechanisms like the cholinergic anti-inflammatory pathway; so in the case of SLE, reduced efferent vagus nerve activity may indirectly affect renal inflammation and therefore hypertension. In this study, we sought to investigate the impact of disrupting vagal neurotransmission on renal inflammation and hypertension in the setting of chronic inflammatory disease. Female SLE (NZBWF1) and control (NZW) mice were subjected to a right unilateral cervical vagotomy or sham surgery and 3 wk later were implanted with indwelling catheters to measure blood pressure. Indices of splenic and renal inflammation, as well as renal injury, were assessed. Unilateral vagotomy blunted SLE-induced increases in mean arterial pressure, albumin excretion rate, and glomerulosclerosis. This protection was associated with reduced splenic T cells and attenuated SLE-induced increases in renal proinflammatory mediators. In summary, these data indicate that unilateral vagotomy reduces renal inflammation and reduces blood pressure in SLE mice. The vagus nerves have myriad functions, and perhaps other neuroimmune interactions compensate for the ligation of one nerve.

Adherence and persistence of mirabegron and anticholinergic therapies in patients with overactive bladder: a real-world claims data analysis

BACKGROUND

Adherence and persistence rates of anticholinergic (ACH) therapies have been well described. To date, few studies describe these metrics for mirabegron in patients with overactive bladder.

METHODS

This retrospective analysis of MarketScan^® database assessed adherence and persistence of patients receiving either mirabegron or ACH. Study eligibility required an index date (first prescription filled) between July 2012 and June 2013 with 12 months of continuous enrolment preindex date and 12 months of follow-up. Adherence was defined as a proportion of days covered of ≥ 80% among patients with at least 2 fills of index medication. Persistence measures included treatment failure described as either treatment discontinuation (medication supply gap ≥ 30 days) or switching to a different medication. A medication supply gap of ≥ 45 days was used as a sensitivity analysis.

RESULTS

The mean age of mirabegron users (n = 4037) was 67 years and 43% were ACH naïve while the mean age of ACH users was 62 years (n = 67,943). Over the 12-month follow-up period, 44% of patients treated with mirabegron and 31% of patients treated with ACH were adherent to their indexed medications. Treatment failure was 81% for mirabegron and 88% for ACH. Most mirabegron treatment failures were because of treatment discontinuation (67%) versus switching to ACH therapy (14%). The ACH discontinuation rate was 84% and treatment switching rate was 4%. The mean (standard deviation) time to treatment failure was 143 (130) days for mirabegron and 69 (69) days for ACH. Adherence and persistence patterns were similar in the sensitivity analysis using a ≥ 45-day supply gap threshold.

CONCLUSIONS

This real-world study demonstrated low adherence and persistence to mirabegron similar to ACH therapies.

Greater fear of hypoglycaemia with premixed insulin than with basal-bolus insulin glargine and glulisine: patient-reported outcomes from a 60-week randomised study

AIM

To assess the effect of initiating insulin treatment on quality of life of patients with type 2 diabetes (T2DM) in the 60-week All-to-Target trial (NCT00384085).

METHODS

Patient-reported outcomes from a phase IV, multicentre, randomised, open-label, parallel-group study were analysed. Participants were randomised to: insulin glargine with up to one insulin glulisine injection (G + 1); insulin glargine with stepwise addition of up to three insulin glulisine injections (G + 3); or twice-daily premixed 70/30 insulin protamine-aspart/aspart (PM-2). Patient-reported outcome questionnaires were administered at weeks 0, 6, 12, 24, 36, 48 and 60.

RESULTS

There were no between-group differences in the Psychosocial Adjustment to Illness State-Self Report (PAIS-SR) or in the EuroQoL Group Five-Dimension Self-Report Index Questionnaire (EQ-5D) from baseline to week 60; however, PAIS-SR scores improved significantly over this period in the G + 3 group (p = 0.0016) and EQ-5D scores worsened significantly in the PM-2 group (p = 0.02). Hypoglycemia Fear Survey Behaviour and Worry subscales worsened significantly for all groups, with greater deterioration being observed in the PM-2 group than in the G + 1 group (Behaviour, p = 0.0050; Worry, p = 0.0017) and G + 3 groups (Behaviour, p = 0.0105; Worry, p = 0.0016). Total scores on the Diabetes Quality of Life (DQoL) questionnaire improved more in the G + 3 group than in the PM-2 group over the study period (p = 0.0284), with all groups showing a significant improvement in DQoL score over time.

CONCLUSION

Insulin glargine-based regimens showed advantages over premixed insulin in a number of patient-reported outcome measures. The potential impact on fear of hypoglycaemia may be of particular relevance when addressing the major barriers to early insulin treatment.

Knowledge, attitudes and barriers regarding vaccination against hepatitis A and B in patients with chronic hepatitis C virus infection: a survey of family medicine and internal medicine physicians in the United States

BACKGROUND

Although vaccination against hepatitis A virus (HAV) and hepatitis B virus (HBV) is recommended for all patients with chronic hepatitis C virus (HCV) infection, physician vaccination practices are suboptimal. Since training for family medicine (FM) and internal medicine (IM) physicians differ, we hypothesised that there are differences in knowledge, attitudes and barriers regarding vaccination against HAV and HBV in patients with chronic HCV between these two groups.

METHODS

A two-page questionnaire was mailed to 3000 primary care (FM and IM) physicians randomly selected from the AMA Physician Masterfile in 2005. The survey included questions about physician demographics, knowledge and attitudes regarding vaccination.

RESULTS

Among the 3000 physicians surveyed, 1209 (42.2%) returned completed surveys. There were no differences between respondents and non-respondents with regard to age, gender, geographic location or specialty. More FM than IM physicians stated that HCV+ patients should not be vaccinated against HAV (23.7% vs. 11.8%, p < 0.001) or HBV (21.9% vs. 10.6%, p < 0.001). FM physicians were also less likely than IM physicians to usually/always test HCV+ patients for immunity against HAV (33.9% vs. 48.6%, p < 0.001) or against HBV (50.8% vs. 68.0%, p < 0.001). There were numerous barriers to HAV and HBV vaccination identified. The median number of barriers was 3 for FM physicians and 2 for IM physicians (p < 0.001).

CONCLUSIONS

Despite recommendations to vaccinate against HAV and HBV in patients with chronic HCV infection, physicians often do not test or vaccinate susceptible individuals. Interventions are needed to overcome the barriers identified and improve vaccination rates.

Ritucharya: Answer to the lifestyle disorders

Ritu, the season, classified by different features expresses different effects on the body as well as the environment. Ayurveda has depicted various rules and regimens (Charya), regarding diet and behavior to acclimatize seasonal enforcement easily without altering body homeostasis. The prime principle of Ayurvedic system of medicine is preventive aspect, can be achieved by the change in diet and practices in response to change in climatic condition. This is a very important aspect of preventive medicine as mentioned in Ayurvedic texts. Lifestyle disorders are very common in the present era, basically originating from lack of following seasonal regimens due to lack of concentration in seasonal characteristics. A firm scientific analysis is the base, which holds true even on date. In this review article, various regimens in diet and lifestyle as mentioned in the classics of Ayurveda and their importance on lifestyle disorders has been discussed.

Exploring feature-based approaches in PET images for predicting cancer treatment outcomes

Accumulating evidence suggests that characteristics of pre-treatment FDG-PET could be used as prognostic factors to predict outcomes in different cancer sites. Current risk analyses are limited to visual assessment or direct uptake value measurements. We are investigating intensity-volume histogram metrics and shape and texture features extracted from PET images to predict patient's response to treatment. These approaches were demonstrated using datasets from cervix and head and neck cancers, where AUC of 0.76 and 1.0 were achieved, respectively. The preliminary results suggest that the proposed approaches could potentially provide better tools and discriminant power for utilizing functional imaging in clinical prognosis.