肺部Ｘ光片
===
###### tags: `Algorithm` `Project`

## AI
> ### Current and emerging artificial intelligence applications in chest imaging: a pediatric perspective
>> https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8409695/
> ### Deep Learning for Chest X-ray Analysis: A Survey
>> https://arxiv.org/pdf/2103.08700.pdf
## Pleural Effusion
> ### Automated Pleural Effusion Detection on Chest X-rays
> https://scholar.smu.edu/cgi/viewcontent.cgi?article=1093&context=datasciencereview
> {%pdf https://scholar.smu.edu/cgi/viewcontent.cgi?article=1093&context=datasciencereview %}

## Pulmonary Nodule
> ### SANet: A Slice-Aware Network for Pulmonary Nodule Detection
> https://mftp.mmcheng.net/Papers/21pami-lungNodule.pdf

## Pneumothorx
> ### Deep learning algorithm for surveillance of pneumothorax after lung biopsy: a multicenter diagnostic cohort study
> https://link.springer.com/article/10.1007/s00330-020-06771-3
> ### Predictors of Pneumothorax/Pneumomediastinum in Mechanically Ventilated COVID-19 Patients
> https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8054543/

## Visualization
> ### A Review of Different Interpretation Methods
> https://mrsalehi.medium.com/a-review-of-different-interpretation-methods-in-deep-learning-part-1-saliency-map-cam-grad-cam-3a34476bc24d

> ### Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
> https://arxiv.org/pdf/1312.6034.pdf
> {%pdf https://arxiv.org/pdf/1312.6034.pdf %}
> ### The false hope of current approaches to explainable artificial intelligence in health care
> https://www.thelancet.com/action/showPdf?pii=S2589-7500%2821%2900208-9
## COVID19
> ### Transfer Learning from CheXNet to COVID-19
> http://cs230.stanford.edu/projects_spring_2020/reports/38949657.pdf

## Competiter
> ### Lunit
> https://asset.fujifilm.com/www/sg/files/2021-05/09d8cad3a2f7c3ded0beee9f66e03200/Lunit_INSIGHT_CXR_Medical_White_Paper.pdf
> ### Applying Advanced Technology in Clinical Practice: Regulatory Approval Cases of AI Software “Lunit INSIGHT
> https://www.imdrf.org/sites/default/files/docs/imdrf/final/meetings/imdrf-meet-190916-russia-yekaterinburg-15.pdf
> ### FDA
> https://www.accessdata.fda.gov/cdrh_docs/pdf21/K211733.pdf
> ### Artificial intelligence for analysing chest X-ray images
> https://www.nice.org.uk/advice/mib292/resources/artificial-intelligence-for-analysing-chest-xray-images-pdf-2285965931918533

## Pricing

| Company / Product | Deploy & Setup Cost | Licence Cost | Unit Cost | Total Cost per Hospital |
| -------- | -------- | -------- | -------- | -------- |
| qXR(Qure.ai) | £6500 (NTD 241,157) | £0.75~£2 (NTD 27.84~74.2) per scan | £0.9 (NTD 33.38) per scan | £37920 (NTD 1,407,050)
| red dot(behold.ai) | £10000 (NTD 371,083) |£60,000 (NTD 2,227,704) | £1.66 (NTD 61.55) per scan | £70000 (NTD 2,597,403)
| Lunit INSIGHT CXR(lunit) | £6,000 (NTD222,770) | £1~£2 (NTD 37.1~74.2) per scan | £1.14 (NTD 42.27) per scan | £48031 (NTD 1,562,227)

> The median number of X-rays done per hospital trust was
42,133 between April 2020 and March 2021 (NHS England 2021)

> Lockwood et al. (2016) outlined a methodology to estimate consultant radiologist hourly unit cost using data from the Personal Social Services Research Unit (PSSRU) Unit Costs of Health and Social Care and other sources. In this study the authors
estimated a unit cost of £156 per hour, equivalent to £2.60 per minute. According to the Royal College of Radiologists (2018) the average time to interpret and report X-ray tests is 2 minutes per image. (This is an isolated component of the reporting activity, and does not reflect the overall time spent per case reported.) Considering such a figure and inflating the cost estimate from Lockwood et al. (2016) to 2019/20 prices using the PSSRU inflation index (Curtis and Burns 2020), the current reporting cost in the NHS would be £11.49 per X-ray image