2.12 Common issues: Server-Side Image-Maps
From the Becta document
Server-side image maps rely on the user being able to use a "point and click" device, such as a mouse. A server-side script records the position (in pixels) of the pointer with respect to the image being used, and triggers the appropriate action (e.g. a hyperlink, for example). Some users are unable to use a mouse, and rely on the use of the keyboard to navigate websites and activate functions within the site. Additionally, it is not possible to have separate alt text for each active region of a server side image map and so visitors without images will not have sufficient alternative text displayed to them.
Client-side image maps (using the MAP and AREA tags in HTML) permit keyboard navigation and activation, and will display more meaningful alternative text and are therefore preferable to the less accessible server-side version. It is also possible to have alternative text for each separate region of the image map.
Very rarely needed. Best practice - Don't use them. See Text-equivalent: Image-map for more details.
9.1 Provide client-side image maps instead of server-side image maps except where the regions cannot be defined with an available geometric shape. [NGfL requirement]