AI Matches Radiologists in Diagnosing Lung Cancer.

 A study published in Radiology found that deep learning artificial intelligence (AI) algorithm diagnosed lung cancer from chest radiographs at a rate similar to radiologists.  The researchers tested it on 10,285 radiographs form 10,202 individuals with 10 radiographs with visible cancer.  The algorithm showed comparable sensitivity 90% to 60% for the radiologists.  In a screening cohort of 100,525 chest x-rays from 50,070 individuals with 47 radiographs with lung cancer, the algorithm's sensitivity was 83% and false positive rate was 3%.  The investigators suggested the algorithm could prove useful especially for clinicians treating healthy persons with lower prevalence of lung cancer.

Adults should start screening for colorectal cancer at age 45

The U.S. Preventive Services Task Force (USPSTF) recent draft recommendation statement proposed that "Adults should start screening for colorectal cancer routinely at age of 45 to 75, instead of waiting until 50 in a move that reflected the sharp rise of the number of colon and rectal cancers in young adults.

Though the vast majority of colorectal cancers are still found in the those of 50 and older, 12 percent of the 147,950 colorectal cancers that will be diagnosed this year, approximately 18,000 cases, will be found in adults under the age of 50.

The American Cancer Society in a statement issued in 2018 recommended starting at 45, after researchers found a sharp rise of the disease among young adults as young as in their 20s and 30s, including a sharp rise in rectal cancers.

The proposal by the USPSTF still must be finalised.   Its guidance on screenings and preventive care services is followed by doctors, insurance companies and policymakers.  

Chest X-rays Highly Predictive of Covid-19

A paper published in Radiology; Cardiothoracic Imaging reports on a study of 366 patients who had a clinical presentation suspicious for Covid-19.  Radiologists reviewed the patients' chest X-rays along with concurrent reverse polymerase chain reaction (RT-PCR)  virus tests.  The patients mean age was 52.7 (range 17-98 years). RT-PCR tests took an average of 2.5 days until a final diagnosis became available. Radiologists found characteristic X-ray pattern such as confluent, ground glass opacification in the mid and lower lung fields highly specific (96.6%) with a positive predictive value of (83.8%) for making the diagnosis of SARS-CoV-2 viral infection.   

Stress Test Cardiac MRI

Cardiac MRI (CMR) is rarely performed in conjunction with stress testing in heart patients, but a large new study may change that. Researchers followed 9,151 patients in real-world settings with known or suspected coronary artery disease (CAD) for up to 10 years following first vasodilator stress CMR. They found the use of stress CMR significantly improved the accuracy of mortality prediction. The study was published online in JAMA Cardiology

Pulmonary Embolism in Patients with Covid-19

A retrospective study published in Radiology: Cardiothoracic Imaging, evaluated 62 Covid-19 patients with computed tomography pulmonary arteriogram (CTPA).

 Thirty-seven percent of patients with confirmed COVID-19 disease were found to have a pulmonary embolism (PE) on CT pulmonary angiography. 

The authors concluded in patients with confirmed Covid-19 almost 40% had pulmonary embolism and PE can be a cause of decompensation.  CTPA can be used to evaluate risk and severity of PE risk on patients with Covid-19 and manage their care.

Brain MRI Findings in Patients with Severe COVID-19

A study published in Radiology found that patients with COVID-19 infection exhibit abnormal brain findings other than stroke on MRI. 

The investigators looked at findings from thirty men and seven women whose mean age was 61 years who met the inclusion criteria. The researchers observed the following; signal abnormalities in the medial temporal lobe were noted in 43% of the patients;  30% of the patients had non-confluent multifocal white matter hyperintense lesions on fluid-attenuated  inversion recovery and diffusion sequences, with variable enhancement, with hemorrhagic lesions (30%); and extensive isolated white matter microhemorrhages (24%). The patients with the worse prognosis were those with brain hemorrhage.

                                   

The researchers concluded that patients with severe Covid-19 who did not have ischemic cerebral infarcts had abnormal MRIs with a wide range of findings.

CT Angiography in Patients with Minor Strokes

A paper published in Radiology reports that CT angiography in patients with acute minor strokes was found to be beneficial in patients who are candidates for thrombectomy. 

The searchers stated that while minor strokes represent up to two-thirds of all cases of acute ischemic strokes, it's unknown whether CT angiography (CTA) to evaluate large-vessel occlusion in patients with minor stroke is cost-effective.

Large vessel occlusion is present in around 18% of patients with National Institutes of Health Stroke Scale (NIHSS) scores of 0-4 and 39% of patients with NIHSS scores of 5-8, but the authors said CTA is only recommended as a cost effective study to triage patients with acute minor stroke and find potential candidates for mechanical thrombectomy.  

The authors concluded that screening for large vessel occlusion with CTA in patients with acute minor stroke is cost effective and is associated with better outcome .

CT Screenings Can Reduce Lung Cancer Mortality

A study published in the New England Journal of Medicine, focused on data from 13,195 men and 2,594 women aged 50 to 74 years old who were heavy smokers, found that participants had a lower mortality if they were screened with computed tomography (CT).

All individuals were followed for a minimum of 10 years and each was randomly assigned to get either low-dose CT scans at baseline, a year, three years and 5.5 years, or no screenings at all. Participants who were screened had lower death rates than the control group: 24% lower for men and 33% lower for women.  The overall referral rate for suspicious nodules was 2.1%.

A U.S. trial of nearly 54,000 heavy smokers in 2011 also found a 20% drop in lung cancer deaths via screenings with CT in comparison to chest X-rays. 

Following the U.S. study, the U.S. Preventive Services Task Force recommended low-dose CT screening for people aged 55 to 80 who had a history of smoking equal to a pack a day for 30 years and were either still smoking or had quit within the past 15 years. 

The American Cancer Society states that lung cancer is responsible for 25% of all cancer deaths and estimates that 72,500 men and 63,220 women in the U.S. are expected to die of lung cancer in 2020

Corona virus; a review and projections of the pandemic.

Since its beginning in December 2019 in Wuhan city in Hubei province of China nearly 860,000 cases and 42,000 deaths due to Corona virus have been reported worldwide as of April 1st, 2020 8:00 GMT.  During this period, confirmed cases in select countries, are as listed in round numbers; US 190,000, Italy 105,000, China 82,000, Germany 71,000, France 53,000, Sweden 4,500, Greece 1,300, S. Korea 9,900 and Japan 2,200.  

During the same time the deaths per million of inhabitants in the above-mentioned countries are; US 12, Italy 200, China 55, Germany 10, France 58, Sweden 18, Greece 5, S. Korea 3 and Japan 0.5.  Interestingly, while China enacted strict quarantine of the 60 million inhabitants of the Hubei province, Japan and Holland lets individuals and businesses continue as in pre Covid-19 time expecting "herd immunity" will deal with the epidemic in a similar way as the annual flu.  S. Korea and Singapore have performed widespread testing of their populations, Sweden instituted social distancing only while the remaining countries enacted lockdown schemes of different severity. 

The first case in Greece appeared on February 25, 2020 and was related to an individual who had travelled to the north of Italy, as well as members a group of pilgrims who had travelled to Israel and Egypt.  

On May 1st the World count of confirmed Corona virus according to Johns Hopkins site was 3,330,000 cases and the death count was 235,000. The U.S. reported 1,100,000 cases and 64,000 deaths.  Next the two most affected countries were Spain with 240,000 cases and 25,000 deaths and Italy with 205,000 cases and 28,000 deaths.  Two countries that have done well are  Germany with 163,000 cases and 6,600 deaths and Austria with 15,500 cases and 585 deaths.  The Greek CDC reported 2,590 cases and 140 deaths.   The majority of those who died (70%) were older than 65 and had pre-existing conditions.  In my opinion, it is the deaths per million of inhabitants, is a useful indicator and points on how and if the countries' health systems cope with the stress of the pandemic.  The deaths per million for select countries are as follows: US 80, Spain 532, Italy 467, France 240, UK 409, France 364,  Sweden 260, U.S. 194, Germany 267, Austria 41 and Greece 14.  S. Korea  has 5, Japan has 3 and Taiwan's has only 0.3 deaths per million.

As of June 1st, 2020, 35,553 patients have been diagnosed with corona virus in the State of Ohio and 2,162 have died, which corresponds to 185 deaths per million.  Ohio and Greece have approximately 10 million inhabitants and have near identical land masses of 50,000 square miles and although geographically distant and with health systems which are distinct, both enacted early lockdowns and both had adequate number of ICU beds fare different when it comes to deaths per million for reasons I cannot explain other than obesity is more prevalent in Ohio. The State with the worst toll is New York with 379,902 cases and 29,918 deaths that corresponds to a death rate of 1,528 deaths per million. 

The Institute of Health Metrics and Evaluation (IHME) of the University of Washington estimates a total of 84,000 deaths for the United States.  Because we know that the deaths per million in Hubei province were 55 per million and because it is the only country that the epidemic has completed its course, I thought we could estimate the death toll in different countries but it turn a grossly inaccurate metric.  As of today, US has 316 deaths per million, UK has 584, Spain has 577, Italy 557, Germany 103, Austria will likely have higher death rates to the projected ones, Germany and Austria 74 and Greece 18.

Four Asian countries have very low death rates and are;  Taiwan with 0.3 deaths per million, Singapore 4, S. Korea 5 and Japan 7.  

Epidemiologists and others are using different models and have and will advance theories why the death rates will differ among countries.  I believe that the estimates of 100-240,000 deaths in the US  are very high while the IHME projection of 84,000 deaths will be closer to the real number at the end of the pandemic.  I also feel Spain's and Italy’s death rate will stay significantly higher than the projected ones while Germany’s, Austria's, Sweden’s, Greece’s, S. Korea’s, Japan’s and especially Taiwan's will be well below projections. In my opinion Greece’s and the State of Ohio death totals will be between 200 and 500 at the end of the pandemic.   I also believe the number of deaths of the United States will be approximately 60,000, which incidentally was the death toll CDC reported for the 2017-2018 flu season.  We will only know the exact number(s) at the completion of the pandemic in August 2020. 

The present was posted at 8:30 GMT on April 1, 2020. I decided I will update the incidence and death numbers  weekly in order for visitors to know the most recent tally of the coronavirus pandemic.

Der Doctor Schnabel Von Rom

Plague doctors were usually inexperienced physicians, hired by towns, who treated patients with bubonic plague.  

A famous doctor whose advice on how to prevent the plague still has some validity was Nostradamus. He adviced the removal of infected corpses, getting fresh air and drinking clean water.  Other famous plague doctors were Giovanni Ventura whom the city of Pavia contracted in 1479 and the Irish physician, Niall O Glacáin (1563-1653) who earned deep respect in Spain, France and Italy for his bravery in treating numerous people with the plague.

In the 17th and 18th centuries, some doctors wore a beak-like masks filled with aromatic items designed to protect them from putrid air, which according to the miasmatic theory of disease was seen as the cause of infections. 

Copper engraving of Doctor Schnabel, circa 1656, by Paulus Fürst.  A plague of a  doctor in 17^th Century Rome with a satirical poem.  The German title translates as Doctor Beak of Rome and depicts a plague doctor in his costume.  The engraving features a satirical Latin/German macaronic poem (‘Vos Creditis, als eine Fabel, / quod scribitur vom Doctor Schnabel’) roughly meaning "a funny tale is disclosed, / by a doctor with a big nose."

Artificial Intelligence (AI) Outperforms Radiologists in Mammography

The aim of screening mammography is to detect breast cancer in women as early as possible before signs of the disease become clinically obvious.  In a study published in Nature McKinney et al found that AI bested radiologists in detecting breast cancer in screening mammograms. 

Mammograms of 25,856 women in the United Kingdom and 3,097 women in the United States were used to train the AI system. AI was then used to identify the presence of breast cancer in mammograms of women who were known to have had either biopsy-proven breast cancer or normal follow-up imaging results at least 365 days later. The study included mammograms; by conventional digital (2D) mammography and tomosynthesis (also known as 3D mammography). 

The authors report that the AI system outperformed diagnoses made by the radiologists who initially interpreted the mammograms, and the decisions of 6 expert radiologists who interpreted 500 randomly selected cases.

The study reports an absolute reduction of 5.7% and 1.2% (USA and UK) in false positives and 9.4% and 2.7% in false negatives. In an independent study of six radiologists, the AI system outperformed all of the human readers: the area under the receiver operating characteristic curve (AUC-ROC) for the AI system was greater than the AUC-ROC for the average radiologist by an absolute margin of 11.5%. The authors also performed a simulation in which the AI system participated in the double-reading process that is common in the UK, and found that the AI system maintained non-inferior performance and reduced the workload of the second reader by 88%. 

The authors suggest that further assessment of the AI system with clinical trials may lead to improvements in the accuracy and efficiency of breast cancer screening by limiting the high rates of false positives and negatives which are known to take place in the interpretation of mammograms.

Adobe on the bay

Adobe on the bay features distinct swaths of white, blue and green, and obscured rainbow moiré patterns. With its jagged, but even edges, it is decidedly removed from any notion of a canvas.  In this digitally manipulated image, my intend was to infer the calmness of our adobe in the little paradise on earth the bay of Porto Rafti, which has a landscape known for its blue sea, blue skies and expansive vistas.

Dear readers please notice the new URL
http://radiologymonthly.blogspot.com/