Validation of clinical acceptability of deep-learning-based automated segmentation of organs-at-risk for head-and-neck radiotherapy treatment planning

Introduction

Organ-at-risk segmentation for head and neck cancer radiation therapy is a complex and time-consuming process (requiring up to 42 individual structure, and may delay start of treatment or even limit access to function-preserving care. Feasibility of using a deep learning (DL) based autosegmentation model to reduce contouring time without compromising contour accuracy is assessed through a blinded randomized trial of radiation oncologists (ROs) using retrospective, de-identified patient data.

Methods

Two head and neck expert ROs used dedicated time to create gold standard (GS) contours on computed tomography (CT) images. 445 CTs were used to train a custom 3D U-Net DL model covering 42 organs-at-risk, with an additional 20 CTs were held out for the randomized trial. For each held-out patient dataset, one of the eight participant ROs was randomly allocated to review and revise the contours produced by the DL model, while another reviewed contours produced by a medical dosimetry assistant (MDA), both blinded to their origin. Time required for MDAs and ROs to contour was recorded, and the unrevised DL contours, as well as the RO-revised contours by the MDAs and DL model were compared to the GS for that patient.

Results

Mean time for initial MDA contouring was 2.3 hours (range 1.6-3.8 hours) and RO-revision took 1.1 hours (range, 0.4-4.4 hours), compared to 0.7 hours (range 0.1-2.0 hours) for the RO-revisions to DL contours. Total time reduced by 76% (95%-Confidence Interval: 65%-88%) and RO-revision time reduced by 35% (95%-CI,-39%-91%). All geometric and dosimetric metrics computed, agreement with GS was equivalent or significantly greater (p<0.05) for RO-revised DL contours compared to the RO-revised MDA contours, including volumetric Dice similarity coefficient (VDSC), surface DSC, added path length, and the 95%-Hausdorff distance. 32 OARs (76%) had mean VDSC greater than 0.8 for the RO-revised DL contours, compared to 20 (48%) for RO-revised MDA contours, and 34 (81%) for the unrevised DL OARs.

Conclusion

DL autosegmentation demonstrated significant time-savings for organ-at-risk contouring while improving agreement with the institutional GS, indicating comparable accuracy of DL model. Integration into the clinical practice with a prospective evaluation is currently underway.

[Dental treatment and oral health care on Tokashiki Island, Okinawa]

Okinawa Prefecture has been promoting dental treatment and oral health care in places where there are no dentists. The Ministry of Health and Welfare has been cooperating with eight dental colleges, including Tokyo Medical and Dental University, for promotion thereof since 1961. The 155th promotion was held on Tokashiki Island. The caries prevalence rate of the students on Tokashiki Island was relatively high compared with the average caries prevalence rate of the same age group throughout Japan, and most of the caries were thought to be due to too many soft drinks. Resin filling was the most popular treatment during this promotion. Severe periodontal disease was observed in middle-aged persons, but we could only perform initial periodontal treatment because of the limited treatment period. Complete or partial dentures were made or repaired for many elderly. The questionnaire study showed most of the denture wearers on Tokashiki Island were satisfied with their dentures, although many dentures did not fit, and the prescription was improved. The period of this promotion was too short to perform complete dental treatment and to prevent caries and periodontal disease, and primary prevention and higher dental hygiene education should be strengthened in no dentist areas like Tokashiki Island.

Long-term durability of dentin bonds made with a self-etching primer, in vivo

The long-term durability of bonds between adhesive resins and dentin is of significant importance for the longevity of bonded restorations. We carried out an in vivo one-year study to evaluate the durability of resin-dentin bonds in the oral cavity, as well as to test the hypothesis that the adhesive interface would show morphological changes in vivo over time. Very shallow saucer-shaped dentin cavities were prepared in 12 intact teeth of one Japanese monkey (Macaca fuscata) under general anesthesia. The cavities were restored with Clearfil Liner Bond II and Clearfil Photo Posterior resin composite. The teeth were extracted at three different times: immediately, and 180 and 360 days after placement of the restorations. One day after the monkey was killed, specimens of the three time periods were subjected to the micro-tensile bond test at a crosshead speed of 1 mm/min. The surfaces of the failed bonds were observed under a field emission scanning electron microscope (FE-SEM). Bond strength measurements in this study were successfully performed and were stable at approximately 19 MPa during the one-year testing. Scanning electron microscopic observations of the failed surfaces revealed, at the top of the hybrid layer and within the adhesive resin, porosity which increased over time. Long-term bonds can be assessed in vivo by the combined evaluation of the microtensile bond strength and SEM morphological examination of the adhesive interface.

Monkey pulpal response to adhesively luted indirect resin composite inlays

Monkey pulpal responses to resin-bonded indirect resin composite inlays were histopathologically evaluated by placing them in either etched-enamel, total-etched, or adhesive-resin-lined cavities. Initial pulpal responses caused by re-exposure of the cut dentin surfaces and luting procedure subsided if a tight marginal seal was secured by final cementation of the inlay. The adhesive resin coating of freshly cut dentinal walls/floors seems to provide a new technique to protect the dentin and pulp in indirect restorations requiring temporary sealing.