Software optical character recognition

If a case report form must be used, the form can be scanned or faxed into a computer system. This form is then read using optical character recognition (OCR) and converted to raw data. The accuracy of OCR applications is high, but the data must be checked to ensure quality. The OCR software can have problems with ambiguous data. For example, the letter 1 and number 1 can look the same on paper. The elimination of the data entry step saves both time and money. Careful design of the case report form and good instructions to the end user will minimize entry errors. 

You can use a scanner and a software program that performs optical character recognition to create a data file from hard copy. Once imported into Excel, the data may have to be manipulated to get it into a useable form. The following example shows how to record a simple Excel macro for converting scarmed data with blank lines and other undesired features into useful columnar data. 

Unfortunately, the range of possibilities for conversion from paper-based graphical format into chemical structures is not so wide. This subject comes close to the graphic-to-text conversion that underpins optical character recognition of a scanned image, and has been termed chemical OCR . The only known programme for enabling this conversion is Kekule supplied by Oxford Molecular (http //www.oxmol. com/software/kekule/spec.shtml), now part of Pharmacopeia (http //www.pcop.com). 

---