Dataset Information

Towards toxic and narcotic medication detection with rotated object detectors.

ABSTRACT:

Background

Recent years have witnessed the advancement of deep learning vision technologies and applications in the medical industry. Intelligent devices for specific medication management could alleviate workload of medical staff by providing assistance services to identify drug specifications and locations.

Methods

In this work, object detectors based on the you only look once (YOLO) algorithm are tailored for toxic and narcotic medication detection tasks in which there are always numerous of arbitrarily oriented small bottles. Specifically, we propose a flexible annotation process that defines a rotated bounding box with a degree ranging from 0° to 90° without worry about the long-short edges. Moreover, a mask-mapping-based non-maximum suppression method has been leveraged to accelerate the post-processing speed and achieve a feasible and efficient medication detector that identifies arbitrarily oriented bounding boxes.

Results

Extensive experiments have demonstrated that rotated YOLO detectors are highly suitable for identifying densely arranged drugs. Six thousand synthetic data and 523 hospital collected images have been taken for training of the network. The mean average precision of the proposed network reaches 0.811 with an inference time of less than 300 ms.

Conclusions

This study provides an accurate and fast drug detection solution for the management of special medications. The proposed rotated YOLO detector outperforms its YOLO counterpart in terms of precision.

SUBMITTER: Peng J

PROVIDER: S-EPMC10102758 | biostudies-literature | 2023 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Towards toxic and narcotic medication detection with rotated object detectors.

Peng Jiao J Wang Feifan F Ma Xiaochi X Chen Zichen Z Fu Zhongqiang Z Hu Yiying Y Zhou Xinghan X Wang Lijun L

Quantitative imaging in medicine and surgery 20230224 4

<h4>Background</h4>Recent years have witnessed the advancement of deep learning vision technologies and applications in the medical industry. Intelligent devices for specific medication management could alleviate workload of medical staff by providing assistance services to identify drug specifications and locations.<h4>Methods</h4>In this work, object detectors based on the you only look once (YOLO) algorithm are tailored for toxic and narcotic medication detection tasks in which there are alwa ...[more]

PMID: 37064387

Similar Datasets

Project description:JUNGFRAU (adJUstiNg Gain detector FoR the Aramis User station) is a two-dimensional hybrid pixel detector for photon science applications at free-electron lasers and synchrotron light sources. The JUNGFRAU 0.4 prototype presented here is specifically geared towards low-noise performance and hence soft X-ray detection. The design, geometry and readout architecture of JUNGFRAU 0.4 correspond to those of other JUNGFRAU pixel detectors, which are charge-integrating detectors with 75 µm × 75 µm pixels. Main characteristics of JUNGFRAU 0.4 are its fixed gain and r.m.s. noise of as low as 27 e(-) electronic noise charge (<100 eV) with no active cooling. The 48 × 48 pixels JUNGFRAU 0.4 prototype can be combined with a charge-sharing suppression mask directly placed on the sensor, which keeps photons from hitting the charge-sharing regions of the pixels. The mask consists of a 150 µm tungsten sheet, in which 28 µm-diameter holes are laser-drilled. The mask is aligned with the pixels. The noise and gain characterization, and single-photon detection as low as 1.2 keV are shown. The performance of JUNGFRAU 0.4 without the mask and also in the charge-sharing suppression configuration (with the mask, with a `software mask' or a `cluster finding' algorithm) is tested, compared and evaluated, in particular with respect to the removal of the charge-sharing contribution in the spectra, the detection efficiency and the photon rate capability. Energy-dispersive and imaging experiments with fluorescence X-ray irradiation from an X-ray tube and a synchrotron light source are successfully demonstrated with an r.m.s. energy resolution of 20% (no mask) and 14% (with the mask) at 1.2 keV and of 5% at 13.3 keV. The performance evaluation of the JUNGFRAU 0.4 prototype suggests that this detection system could be the starting point for a future detector development effort for either applications in the soft X-ray energy regime or for an energy-dispersive detection system.

Project description:BackgroundOvertreatment with cardiometabolic medication in older patients can lead to major adverse events. Timely deprescribing of these medications is therefore essential. Self-reported willingness to stop medication is usually high among older people, still overtreatment with cardiometabolic medication is common and deprescribing is rarely initiated. An important barrier for deprescribing reported by general practitioners is the patients' unwillingness to stop the medication. More insights are needed into the influence of patients' characteristics on their attitudes towards deprescribing and differences in these attitudes between cardiometabolic medication groups.MethodsA survey in older people using cardiometabolic medication using the revised Patients' Attitudes Towards Deprescribing (rPATD) questionnaire was performed. Participants completed the general rPATD and an adapted version for four medication groups. Linear and ordinal logistic regression were used to assess the influence of age, sex, therapeutic area and number of medications used on the patients' general attitudes towards deprescribing. Univariate analysis was used to compare differences in deprescribing attitudes towards sulfonylureas, insulins, antihypertensive medication and statins.ResultsOverall, 314 out of 1143 invited participants completed the survey (median age 76 years, 54% female). Most participants (80%) were satisfied with their medication and willing to stop medications if their doctor said it was possible (88%). Age, sex and therapeutic area had no influence on the general attitudes towards deprescribing. Taking more than ten medicines was significantly associated with a higher perceived medication burden. Antihypertensive medication and insulin were considered more appropriate than statins, and insulin was considered more appropriate than sulfonylureas not favouring deprescribing.ConclusionsThe majority of older people using cardiometabolic medication are willing to stop one of their medicines if their doctor said it was possible. Health care providers should take into account that patients perceive some of their medication as more appropriate than other medication when discussing deprescribing.

Dataset Information

Towards toxic and narcotic medication detection with rotated object detectors.

Background

Methods

Results

Conclusions

Publications

Towards toxic and narcotic medication detection with rotated object detectors.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets